This article explains the behavioral biases that most commonly damage trading results. It is educational and general, a description of how these biases work and how a rules-based process is built to neutralise them, not personalised investment advice. Nothing here is a recommendation to buy or sell any security, and describing a bias is not a claim that avoiding it guarantees profit.
Most trading losses are not caused by bad analysis. They are caused by good analysis overridden, at exactly the wrong moment, by a human being who is frightened, hopeful, or bored. Decades of research in behavioral economics have catalogued the specific, repeatable errors that do this: we sell winners too early and cling to losers, we feel a loss far more sharply than an equal gain, we extrapolate the recent past, and we chase what everyone else is buying. Here is what those biases are, why each one degrades trading results, and why removing the human from the decision is the entire point of a mechanical, rules-based system.
The short version
Behavioral biases are systematic, predictable errors in judgement that arise from the mental shortcuts humans use under uncertainty. In trading they surface as the disposition effect (cutting winners, riding losers), loss aversion, recency bias, and FOMO. A rules-based process counters them by deciding in advance and removing discretion at the moment of action.
Where the biases come from
The modern study of these errors begins with Daniel Kahneman and Amos Tversky, whose work in the 1970s and 1980s showed that people do not evaluate risky choices the way classical economics assumed. Their prospect theory (1979) established two findings that sit underneath almost every trading mistake in this article. First, people judge outcomes as gains and losses relative to a reference point, usually their entry price, rather than in terms of final wealth. Second, they showed losses loom larger than equivalent gains, a phenomenon they named loss aversion; their later work put the asymmetry at roughly twice, a ratio widely reproduced since. Building on Richard Thaler’s work on mental accounting, Hersh Shefrin and Meir Statman later named the disposition effect to describe the tendency to realise gains too readily and hold losses too long.
The important word is systematic. A random mistake averages out over many trades; a bias does not, because it pushes every decision in the same direction. That is what makes these errors dangerous to a trader and, at the same time, what makes them tractable for a rule: a predictable error can be predicted, and a predictable error can be designed around.
The biases that do the most damage
The disposition effect: cutting winners, riding losers
This is the single most documented trading error, and it is exactly backwards from what a sound strategy requires. Under loss aversion, a position showing a profit is a gain the trader is desperate to lock in before it evaporates, so winners are sold early. A position showing a loss is a loss the trader refuses to accept as final, so it is held in the hope it “comes back.” The result is a book of small realised gains and large unrealised losses, the precise opposite of the “cut losses short, let winners run” discipline that every trend-following and momentum approach depends on. Shefrin and Statman named the pattern, building on Thaler’s mental-accounting framing, and later Terrance Odean’s study of real brokerage accounts found it remarkably consistent across ordinary investors.
Loss aversion: the asymmetry underneath everything
Loss aversion is the engine that drives the disposition effect and much else. Because a loss is felt about twice as intensely as an equal gain, the emotional stakes of any bet are lopsided for the trader. This produces two opposite failures depending on the day. It makes people take too little risk when they should stay in a working trade (grabbing a small profit to avoid the possibility of giving it back), and too much risk when they are already losing (holding or even adding to a losing position to avoid crystallising the pain). Neither reaction has anything to do with the merits of the position.
Recency bias: the last few trades feel like the truth
Recency bias is the tendency to weight the most recent events far more heavily than the longer record when forming an expectation. A trader who has just had three winners feels invincible and sizes up; one who has just had three losers feels the strategy is “broken” and abandons it, often right before it works again. Because any real edge is a statistical tendency that plays out over many trades, judging it by the last handful is guaranteed to mislead. Recency bias is why people chase whatever has worked lately and capitulate on whatever has not, both at the worst possible time.
Anchoring: getting stuck on an irrelevant number
Anchoring is the tendency to fixate on a reference number, often the first one encountered, and judge everything relative to it even when it is irrelevant. In trading the classic anchor is the purchase price: a stock “should” get back to what the trader paid, as if the market cared what they paid. Other anchors are a prior high, a round number, or an analyst target. Each turns a fresh decision (is this position worth holding on its current evidence?) into a comparison against a number that carries no information about what happens next.
Overconfidence: mistaking a good run for skill
Overconfidence is the well-documented tendency to overestimate the accuracy of one’s own judgement. In markets it shows up as trading too often, sizing too large, and under-appreciating how much of a result was luck. Odean’s research linked overconfidence directly to excessive trading and lower net returns. It is particularly dangerous after a winning streak, when recency bias and overconfidence reinforce each other, and the trader concludes a lucky run was earned skill just before mean reversion collects the bill.
Confirmation bias: only seeing the supporting evidence
Confirmation bias is the habit of seeking, noticing, and believing information that supports a position already held, while discounting anything that contradicts it. A trader long a name reads the bullish commentary and dismisses the bearish, then feels more certain than the evidence warrants. It is what keeps a losing thesis alive long after the facts have turned, because the trader is curating a one-sided view of reality rather than testing the position against it.
FOMO: buying because everyone else already did
The fear of missing out is the emotional pressure to enter a move that is already well underway, driven by watching others profit and dreading being left behind. FOMO is how traders buy at the top: no setup is involved, only the pain of watching a rocket leave without them, which is precisely when the risk is highest and the remaining upside lowest. It is the disposition effect’s mirror image on the entry side, an emotion, not an edge, deciding when to act.
At a glance: the bias, the tell, and the rule that counters it
The biases are different, but the antidote has one shape: decide the rule in advance, when calm, and let it act at the moment when the emotion would otherwise take over. The table pairs each bias with how it shows up at the desk and the kind of mechanical rule designed to neutralise it.
| Bias | How it shows up in trading | How a rule counters it |
|---|---|---|
| Disposition effect | Selling winners early, holding losers hoping they recover | Predefined exits: a mechanical trailing rule lets winners run and a fixed invalidation closes losers, regardless of feeling |
| Loss aversion | Too little risk on winners, too much on losers | Risk-first, rule-based sizing set before entry, so the loss is bounded and pre-accepted rather than negotiated live |
| Recency bias | Sizing up after wins, abandoning the plan after losses | A fixed process applied every day, so the last few trades do not change how the next one is taken |
| Anchoring | Waiting for a stock to “get back to” the purchase price | Decisions keyed to current evidence and predefined levels, not to the entry price the market cannot see |
| Overconfidence | Trading too often, sizing too large after a good run | A fixed candidate bar and consistent sizing that a winning streak cannot inflate |
| Confirmation bias | Only reading evidence that supports the open position | A rule that evaluates the same measurable inputs for every name, with no narrative to defend |
| FOMO | Chasing a move that has already run, buying at the top | Entries defined by a setup and filled on a rule, so an extended, already-run name ranks down instead of getting chased |
Why rules, not willpower
The natural response to a list of biases is to resolve to try harder, to be more disciplined, less emotional, more objective. This almost never works, and the research explains why: these biases are features of how the mind processes uncertainty and loss. They operate fastest under exactly the conditions trading creates, real money at risk, time pressure, and an outcome that feels personal. Knowing about a bias does not switch it off. Kahneman himself was candid that a lifetime of studying these errors did not make him immune to them.
The workable answer is structural, not motivational: take the decision out of the moment. A mechanical, rules-based process decides in advance, when no position is open and no money is moving, what will constitute a candidate, how much risk a position may carry, and the exact conditions under which it will be exited. Then, at the moment of action, the moment when fear, hope, and the crowd would otherwise take over, there is nothing left to decide. The rule already decided. This is the core reason systematic strategies exist, and it is the same argument behind whether stock signals actually work: a signal’s real value is a precommitment that removes the discretionary override where most of the damage is done.
How Shishin removes the discretion
Removing the human from these decisions is the entire point of a systematic system. Shishin is a rules-based, non-advisory signal service, and each of the biases above maps onto a design choice that leaves nothing to nerve. Names earn a place on the daily ranked board by clearing the same measurable bar every day, which is a defence against recency bias, overconfidence, and confirmation bias at once: a hot streak cannot inflate the bar, and there is no open-position narrative for the model to defend. Risk is set first, before conviction, so loss aversion never gets to renegotiate a position’s size once it is losing, an approach explored in risk per trade. And exits are rules, not moods: a mechanical trailing discipline is designed to let winners run rather than grabbing a quick profit, while a predefined invalidation closes losers rather than nursing them, which is precisely the disposition effect turned on its head.
The regime-routed guardians, Genbu, Suzaku, Byakko and Seiryū, each act only in the market state they suit, which is itself an anti-FOMO and anti-recency mechanism: the system acts when its rules and the environment agree, and it stands aside otherwise. Because the process is fixed and the record is meant to be inspected rather than taken on trust, the whole of it is published and independently attested at the verification log. The point is that a rule cannot panic, cannot fall in love with a loser, and cannot feel the fear of missing out.
The limits: rules move the problem, they do not delete it
Honesty requires the caveat. A rules-based process relocates behavioral risk. The discretion is removed from the individual trade, but two human decisions remain. The first is designing the rules, and biases can be baked into a system just as easily as into a discretionary call: a backtest curve-fitted to the past is overconfidence in numerical form, which is why backtests can mislead and why survivorship and other traps have to be controlled for. The second, and harder, is the discipline to follow the rules in a drawdown, precisely when recency bias screams that the system is broken and the temptation to override it is strongest. A rule only protects a trader who lets it. What a good systematic process buys is a structure that makes the right behaviour the default and the emotional override a deliberate, visible act rather than a reflex.
So: which biases wreck traders, and how do rules avoid them?
The most destructive are the disposition effect (cutting winners, riding losers), loss aversion (the asymmetry that drives it), recency bias, anchoring, overconfidence, confirmation bias, and FOMO. They are systematic, not random, so they do not average away, and they operate fastest under exactly the pressure that trading creates. Willpower is a poor defence because these are features of the mind, not lapses of it. The durable answer is to decide in advance and remove discretion at the moment of action, which is what a mechanical, rules-based process is for. Nobody becomes immune. The disciplined choice simply becomes the automatic one, and the emotional override becomes something a trader has to do on purpose.
Sources & further reading
- Kahneman, D. & Tversky, A. (1979). “Prospect Theory: An Analysis of Decision under Risk.” Econometrica, 47(2), 263 to 291., loss aversion and the reference-point framing that underlies most trading biases.
- Tversky, A. & Kahneman, D. (1992). “Advances in Prospect Theory: Cumulative Representation of Uncertainty.” Journal of Risk and Uncertainty, 5(4), 297 to 323., the later estimate that put loss aversion at roughly twice.
- Shefrin, H. & Statman, M. (1985). “The Disposition to Sell Winners Too Early and Ride Losers Too Long.” Journal of Finance, 40(3), 777 to 790., naming the disposition effect, building on Thaler’s work.
- Odean, T. (1998). “Are Investors Reluctant to Realize Their Losses?” Journal of Finance, 53(5), 1775 to 1798., the disposition effect and overconfidence in real brokerage accounts.
- Kahneman, D. Thinking, Fast and Slow (2011)., a general-audience synthesis of anchoring, recency, overconfidence and the two-system account of judgement.
- Related reading on turning discipline into rules: do stock signals work, risk per trade, and position sizing by conviction.