Skip to main content
What Does Not Work, and the Data

Confirmation Bias in Charting

Pomegra Learn

How Traders See What They Want to See on Charts

Confirmation bias in charting is the tendency to search for, interpret, and recall information in ways that confirm a pre-existing belief or position. A trader who is bullish on a stock looks at a chart and sees bullish patterns—breakouts, higher lows, volume confirmation. A trader who is bearish on the same chart looks at the same data and sees bearish patterns—resistance, lower highs, bearish divergences. Both traders are looking at identical price data, but they are seeing different patterns because they are selectively attending to evidence that supports their existing view.

This bias is insidious because it operates unconsciously. A trader doesn't deliberately ignore bearish signals; they genuinely don't notice them or interpret them away. A lower high is seen as "consolidation before the next leg up" rather than a reversal signal. A breakdown of support is seen as a "bull trap" rather than a genuine decline. The trader's brain is performing a sophisticated feat of motivated reasoning, protecting the existing position by filtering evidence.

Confirmation bias causes traders to interpret the same chart data differently depending on their existing beliefs. A trader bullish on a stock sees bullish patterns and ignores bearish ones; a bearish trader sees the opposite. Selective attention costs real money when it prevents traders from acknowledging when they're wrong.

Key takeaways​

  • Confirmation bias is the selective seeking and interpretation of evidence to confirm existing beliefs; it's automatic and mostly unconscious.
  • In charting, confirmation bias leads traders to see patterns that support their view (bullish breakout if they're long, bearish breakdown if they're short) and ignore contradictory patterns.
  • Traders are especially susceptible when they have an open position; a trader long 1,000 shares will interpret ambiguous signals more bullishly than a trader with no position.
  • The cost of confirmation bias is staying in losing trades too long, averaging down into losing positions, and ignoring stop-loss signals.
  • The solution is systematic rules, pre-commitment to exits, and outsourcing chart interpretation to algorithms or a neutral observer.

The Mechanism: How Confirmation Bias Works in Trading​

Confirmation bias operates through several mechanisms in trading:

Selective exposure: A trader who believes a stock is going up seeks out bullish news and analysis while avoiding bearish sources. They might follow bullish analysts on Twitter and unfollow bearish ones. They might read a bullish article and ignore a bearish article about the same stock. Over time, their information diet becomes skewed toward supporting their position.

Selective interpretation: When a trader encounters ambiguous information, they interpret it in a way that supports their existing view. A stock that closes up 0.5% on a day with heavy selling is interpreted as "resilient and found support" (bullish) by a long trader and "weak, couldn't hold gains" (bearish) by a short trader. The data is the same; the interpretation differs.

Selective recall: A trader remembers the trades that confirmed their beliefs and forgets or downplays trades that contradicted them. A trader might remember the three times a particular pattern led to a winning trade and forget the seven times it led to a loss.

Selective search: When faced with a decision, a trader seeks out evidence supporting their preferred outcome. If they want to stay in a trade, they search for reasons to stay and ignore reasons to exit. If they want to buy a stock, they seek out bullish signals and ignore bearish ones.

These four mechanisms reinforce each other and create a self-reinforcing feedback loop. The trader becomes increasingly convinced of their view because all the evidence they're attending to supports it, while contradictory evidence is filtered out.

The Role of Position Bias​

Confirmation bias is amplified when a trader has an open position. The sunk-cost fallacy—the tendency to continue investing in something because of previous investment—combines with confirmation bias to create strong resistance to admitting a mistake.

A trader who bought a stock at 50 and has watched it decline to 45 is now holding a losing position. When they look at the chart, the bias pushes them toward seeing reasons to hold:

  • "The decline is just a pullback; the stock will recover."
  • "The volume is decreasing, which is often a sign of capitulation."
  • "The stock found support here; I'm confident it will bounce."

A trader with no position, looking at the same chart, might see:

  • "The stock broke below support; this is a bearish breakdown."
  • "The declining volume suggests weak buying interest."
  • "Support is tested, but it's not holding. Could be headed lower."

The difference is position bias: traders unconsciously interpret chart evidence more favorably when they have money at stake. This is not malicious; it's a defense mechanism. But it costs real money when the unfavorable interpretation was correct, and the trader's confirmation-biased view caused them to hold too long.

Studies confirm this: Mehra and Sah (2002) found that traders with open positions were more likely to interpret new information as supporting their position, and they were slower to acknowledge losses than traders without positions. Odean (1998) found that traders held losing stocks longer than winning stocks, partly due to the disposition effect (selling winners early and holding losers) and partly due to confirmation bias (seeing hope in losing stocks while looking for reasons to sell winning ones).

Confirmation Bias in Pattern Recognition​

Confirmation bias is especially dangerous when applied to chart patterns because patterns are inherently subjective. Many possible patterns can be drawn on any chart. A trader can find a head-and-shoulders pattern if they look for one, or a triangle, or a double bottom. They see the pattern they're looking for.

Example: A stock trades in a range from 100 to 110 for three months. A trader who believes the stock is heading higher looks at this and sees an "accumulation pattern" or "consolidation before breakout." They interpret the upper bound as resistance and the lower bound as support, expecting a breakout above resistance. A trader who believes the stock is heading lower looks at the same range and sees a "distribution pattern" or "resistance formation." They interpret the upper bound as trapped buyers and the lower bound as support, expecting a breakdown.

Both interpretations are plausible; the pattern truly is ambiguous. But the trader's pre-existing bias determines which interpretation they select. If the stock breaks upward, the bullish trader is validated ("I called the breakout"). If it breaks downward, the bearish trader is validated ("I called the breakdown"). Whichever one was right claims pattern-reading skill; both were really just benefiting from a 50/50 outcome and confirmation bias.

Bulkowski's (2000) statistical analysis of chart patterns addressed this problem directly. He looked at thousands of real chart patterns and calculated success rates. The results: most chart patterns are only slightly more predictive than random. An average head-and-shoulders pattern has about a 65% success rate, double tops about 60%, triangles about 60–70% depending on type. These are barely better than a coin flip (50%). The reason they look better is that traders remember the successful patterns and forget the failed ones—a classic confirmation bias artifact.

Flowchart​

The "I Was Right, I Was Just Early" Trap​

Confirmation bias often manifests as the "I was right, I was just early" narrative. A trader enters a trade expecting a move in one direction, but the market moves in the opposite direction. Rather than acknowledging the trade was wrong, the trader reinterprets the move as "a delay" or "a pullback before the expected move."

Example: A trader goes long crude oil at 80 expecting a move to 100 based on a chart pattern and supply-demand dynamics. Over the next three months, crude oil declines to 70. The trader's original forecast was not wrong, they tell themselves; the market is just giving a better entry. They average down, buying more at 75. Crude oil continues declining to 60, and the trader is now down 20% on their position.

This narrative—"I was right, just early"—is a form of confirmation bias. The trader is selectively interpreting the price decline not as evidence against their thesis but as a temporary setback. They search for reasons to believe the thesis is still valid ("OPEC cuts production next month") while ignoring reasons to exit ("Demand is weak, supply is strong").

In reality, sometimes traders are right but early; it happens. The issue is that confirmation bias makes this excuse too easy to use. Traders use it to justify staying in losing positions that should be exited. A more rational approach is: "My thesis is still valid if X happens by Y date. If X doesn't happen, I'm wrong, and I exit." This pre-commitment makes it harder to rationalize away losses.

Ambiguous Signals and the Selective Interpretation Game​

Many technical signals are ambiguous and subject to interpretation. The same signal can be read as bullish or bearish depending on context:

  • Volume decline: Could indicate capitulation (bullish reversal) or weakening buying interest (bearish continuation).
  • Reversal candle: Could signal a trend reversal or a false breakout shakeout.
  • Higher high, lower volume: Could indicate declining momentum (bearish) or accumulation (bullish).
  • Lower high, higher volume: Could indicate selling pressure (bearish) or accumulation after a shakeout (bullish).

A trader with a position interprets ambiguous signals in a way that supports their position. A trader long a stock interprets higher volume on down days as "profit-taking," not "selling pressure." A trader short a stock interprets higher volume on up days as "short covering," not "buying strength." Both traders are engaging in selective interpretation of the same data.

The solution to this problem is to use clear, mechanical rules for interpretation. Instead of "interpret volume as you see fit," a rule might be: "If volume is above the 20-day average on down days for three consecutive days, the signal is bearish." This removes discretion and the opportunity for confirmation bias to distort interpretation.

The Availability Bias Compounding Effect​

Confirmation bias is often compounded by availability bias—the tendency to overweight evidence that comes easily to mind. A trader who recently read a bullish article on a stock or who heard a bullish analyst on financial media will overweight that information when looking at the chart. The information is "available" in memory, so it disproportionately influences interpretation.

A study by Tversky and Kahneman (1973) on the availability heuristic showed that people estimate the probability of an event based on how readily examples come to mind. A trader who recently saw three bullish articles about Apple will overestimate the probability of Apple going up, relative to a trader who hasn't been exposed to that information. When they look at Apple's chart, they interpret ambiguous signals more bullishly.

Media and social platforms amplify this effect. A trader spending time on financial Twitter will be exposed disproportionately to bullish or bearish sentiment depending on which analysts they follow. If they follow mostly bullish analysts, they'll develop a bullish bias that gets reinforced by confirmation bias when they look at charts.

Real-world examples​

The Dot-Com Believers (1999–2000): Traders and analysts who believed in the internet revolution in 1999 looked at soaring tech stocks and saw evidence of a new paradigm. Charts showed vertical moves with increasing volume. They interpreted this as "real growth momentum" and "the new economy." Traders who were skeptical looked at the same charts and saw "speculative bubble" and "unsustainable valuations." Both groups were looking at the same data; confirmation bias determined their interpretation. The bubble popped in 2000–2002, but before it did, confirmation bias kept believers bullish for far longer than was rational.

Tesla Bulls vs. Bears (2020–2021): Tesla stock became a Rorschach test for confirmation bias. Bulls saw a chart showing an unstoppable uptrend, revolutionary technology, and exponential adoption. Bears saw the same chart and saw parabolic growth heading toward a crash, an overvalued company, and hype. Both groups had chart evidence supporting their view, because they were selectively interpreting ambiguous signals (declining volume on the way up, large ranges indicating uncertainty). The stock rose to 900 in late 2021, then crashed to 100 in 2023. Both bulls and bears could point to moments when "they were right," because they had selectively remembered supporting evidence and forgotten disconfirming evidence.

Long-Term Capital Management (1998): Before its collapse, LTCM managers looked at their models and the market data and saw confirmation of their thesis that certain spreads were mispriced. When the spreads widened against them, they interpreted the move not as evidence against their thesis but as the market creating an even better opportunity. They averaged down (buying more bonds expecting the spread to tighten). The spread widened further, and LTCM lost billions. Confirmation bias led them to misinterpret evidence against their position as evidence for it.

Common mistakes​

  • Staying in a losing trade while telling yourself you're "just early": Pre-commit to an exit rule before entering. If you say "I exit if the stock closes below 45 for 3 consecutive days," then exit when that rule is triggered, rather than reinterpreting the decline.
  • Selectively reviewing chart patterns that worked while ignoring those that didn't: Keep a log of all the chart patterns you identify and their outcomes, including the failed ones. This forces you to confront the true success rate rather than relying on selective recall.
  • Following analysts or media sources that align with your view: Deliberately read bearish analysis if you're bullish and bullish analysis if you're bearish. Seek out disconfirming evidence.
  • Interpreting ambiguous signals in whatever direction supports your position: Use clear rules for signal interpretation. If a signal can be read two ways, don't rely on it.
  • Averaging down into a losing position: Averaging down amplifies confirmation bias; the trader becomes more committed to the original thesis and more likely to dismiss contradictory evidence.

FAQ​

How can I identify when I'm subject to confirmation bias on a chart?​

Ask yourself: "Would I interpret this signal the same way if I had the opposite position?" If the answer is no, confirmation bias is likely present. Also ask: "What evidence would convince me I'm wrong?" If you can't answer clearly, you're probably using confirmation bias to avoid confronting the evidence against your position.

Should I avoid reading bullish/bearish analysis that aligns with my view?​

You can read it, but deliberately seek out the opposite view as well. If you're long a stock, read at least one bearish analysis per week. If you're short, read a bullish one. This helps counteract the availability bias and confirmation bias that naturally push you toward supporting information.

Can confirmation bias be completely eliminated?​

No, but it can be managed. The most effective management technique is using mechanical rules for entries and exits rather than relying on chart interpretation. If you must use chart interpretation, get a second opinion from someone without a position, or use a checklist of signals that must align before you trade.

Is confirmation bias stronger for day traders or long-term investors?​

It's probably stronger for long-term investors because they have more time to rationalize a position and more opportunity to use reinterpretation to avoid admitting the position was wrong. Day traders face more immediate market feedback (profits and losses within hours) and less time for motivated reasoning. However, short-term traders can also be subject to it within their timeframe.

How does confirmation bias differ from the disposition effect?​

The disposition effect is the tendency to sell winning trades early and hold losing trades too long. Confirmation bias is the tendency to selectively interpret evidence to support an existing position. They often occur together: a trader holds a losing stock partly due to the disposition effect (aversion to losses) and partly due to confirmation bias (selective reinterpretation of negative signals as temporary).

Should I use automated trading systems to avoid confirmation bias?​

Automated systems can help by removing discretion, but they have their own biases. If you build an automated system based on a pattern you believe in, you're likely to overfit it to historical data due to confirmation bias about the pattern. The solution is rigorous out-of-sample testing and live trading validation.

What's the role of personality in confirmation bias?​

Some people are more prone to confirmation bias than others. Those with higher need for closure (preferring definite answers) and those with strong beliefs are more susceptible. These traits are relatively stable, but confirmation bias can be mitigated through deliberate practices: soliciting opposing viewpoints, using clear rules, and regularly confronting contradictory evidence.

Summary​

Confirmation bias in charting leads traders to selectively seek, interpret, and recall evidence that supports their existing position while ignoring or reinterpreting contradictory evidence. The bias is automatic and mostly unconscious, making it difficult to recognize in oneself. It is amplified by position bias (traders with open positions are more susceptible) and availability bias (recent information disproportionately influences interpretation). The cost of confirmation bias is staying in losing trades too long, averaging down into deteriorating positions, and making repeated errors that could have been prevented by acknowledging contradictory evidence early. The solution is systematic rules, pre-commitment to exits, deliberate exposure to opposing viewpoints, and outsourcing interpretation to algorithms or neutral observers. Traders who effectively manage confirmation bias exit losing positions faster and experience lower drawdowns than those who don't.

Next​

→ The Cost of Trading Too Much