Core Behavioural Biases for the NEPSE Investor

First published 23 Aug 2026 · Last verified 29 Aug 2026

Lesson 50 left the Nepali investor with an uncomfortable diagnosis: systematic underperformance is not mostly bad luck or bad regulation — it is bad decision-making, repeated at scale, across hundreds of thousands of BOID (Beneficiary Owner Identification) accounts. This chapter opens the toolbox that explains why. Behavioural finance is the branch of economics that studies how real human beings — not the perfectly rational "economic man" of textbooks — actually make decisions involving money, risk, and uncertainty. Its central finding, replicated across dozens of markets and confirmed by empirical studies of NEPSE investors themselves, is that the human brain uses mental shortcuts, called heuristics, to cope with complexity. These shortcuts work well enough in daily life — deciding which vegetable vendor at the Kalimati market is trustworthy, or judging whether a monsoon cloud means you should turn back on the trek. But in a modern, fast-moving, emotionally charged stock market, the same shortcuts systematically distort judgment. They are not signs of stupidity. They are the predictable output of a normal brain applying everyday reasoning to an environment it did not evolve for. Every investor has these biases. The professional investor's edge is not the absence of bias — it is the discipline of recognising it and building rules that route around it.

This chapter examines eight biases that recur constantly in the Nepali retail investing story: herd behaviour, anchoring, confirmation bias, loss aversion paired with its market cousin the disposition effect, overconfidence, recency bias, FOMO (fear of missing out), gambler's fallacy, and home bias. Each is explained mechanically — what is happening inside the mind — and then grounded in a scenario that will feel familiar to anyone who has watched a Viber investment group light up during a hydropower rally, or has held a loss-making microfinance script for three years "because it will come back." Understanding these patterns will not make them disappear. But naming a bias while it is happening — recognising the specific mental trap you are standing in — is the single most effective tool available for interrupting it before it costs you money.

Lesson 51.1 — Herd Behaviour: The Psychology of the Crowd

Herd behaviour, or herding, describes the tendency of individuals to align their decisions with the actions of a larger group, rather than relying on their own independent analysis. The mechanism is partly informational and partly social. Informationally, an investor reasons: "Surely all these other people buying this hydropower script know something I don't — why would so many people be wrong?" This is a reasonable inference in many walks of life. If a long queue forms outside a momo shop in New Road that you have never tried, the queue itself is useful evidence that the food is good. But financial markets are different from momo queues in one crucial way: the queue outside the momo shop does not change the quality of the momos, whereas a queue of buyers in a stock does change the price you must pay, and a rising price does not by itself mean the underlying business has become more valuable. Socially, herding is also driven by a fear of regret that is asymmetric: an investor who loses money alongside everyone else feels less foolish than an investor who loses money by going against the crowd. Being wrong in company feels safer than being wrong alone, even though the financial outcome is identical.

NEPSE's structure makes herding unusually powerful compared to more institutionally dominated markets. Retail investors are estimated to account for the overwhelming majority of daily turnover on the exchange, and a large share of that retail base coordinates informally through Viber groups, Telegram channels, Facebook pages, and YouTube "market analysis" channels that often number in the tens of thousands of members. When one of these channels flags a script — "Buy XYZ Hydropower before it hits circuit tomorrow!" — thousands of members can place near-identical buy orders within minutes of each other, because the Trading Management System (TMS) used by NEPSE brokers makes order placement fast and frictionless from a mobile phone. The effect is self-reinforcing: as the herd buys, the price rises; the rising price becomes proof, in the herd's own mind, that the tip was correct; the rise attracts a second wave of buyers who were not in the original channel but see the price move on their trading app and chase it. This is precisely the mechanism academic researchers have documented in Nepalese stock market studies — herding intensifies specifically during periods of high market-wide movement, because uncertainty is highest exactly when the pull toward safety-in-numbers is strongest.

Consider the pattern that has repeated across multiple NEPSE cycles: a sector — hydropower in one season, microfinance and finance companies in another, insurance in a third — becomes the "hot" sector. Prices of a handful of scripts in that sector begin moving up on genuine news (a new IPO subscription record, a favourable regulatory circular, a strong quarterly report). Social media commentary amplifies the move. Within weeks, scripts across the entire sector are rising together regardless of whether the individual companies share the same fundamentals — a well-run hydropower company with contracted power purchase agreements and a speculative, under-construction project with cost overruns can trade at similarly inflated valuations, simply because both wear the same sector label the herd is chasing. When the rally eventually runs out of new buyers — which it always does, because no sector can absorb unlimited capital forever — the same herd mentality reverses violently. Sell orders cascade because everyone is trying to exit through the same narrow door at once, and because in a market with daily price bands (circuit breakers), a stock in freefall may not even find a buyer at the lower limit, leaving sellers stuck.

The practical cost of herding is that it forces you into the market's worst possible entry and exit timing. By definition, you can only follow a herd after it has already started moving — which means the herd investor systematically buys after a meaningful part of the up-move has already happened, and sells after a meaningful part of the down-move has already happened. Over many cycles, this converts what should be a value-driven return into a "buy high, sell low" pattern — the exact opposite of profitable investing. The antidote, developed further in later chapters, is not to ignore the crowd's existence but to treat crowd enthusiasm as a signal to slow down and re-verify your own reasoning, rather than a signal to hurry up.

Lesson 51.2 — Anchoring and the Ghost of the Previous High

Anchoring bias is the tendency to rely too heavily on a piece of information encountered early — the "anchor" — when making subsequent judgments, even when that information has little or no logical connection to the correct answer. The anchor does not have to be relevant to be powerful; it simply has to be the first number that lodges in the mind. In investing, the two most common anchors are a stock's previous all-time high price and an investor's own purchase price, and both distort decision-making in remarkably similar ways even though they arise from completely different sources.

Take the previous-high anchor first. Suppose a hydropower script traded as high as Rs 850 during a bull phase eighteen months ago and has since fallen to Rs 340 as the sector cooled and the company's actual earnings disappointed relative to the hype. An investor evaluating this stock today will very often frame the decision not in terms of "is Rs 340 a fair price for this company's future cash flows" but in terms of "this stock used to be worth Rs 850, so at Rs 340 it is cheap, it has 'more room to recover.'" This reasoning treats the old high as a kind of gravitational centre the price ought to return to, when in fact the Rs 850 price may itself have been a herd-driven overvaluation that had nothing to do with the company's fundamentals, and the business today may genuinely be worth less than Rs 340 if its earnings power has permanently deteriorated. The number Rs 850 is psychologically vivid — it sits in the investor's memory, in old screenshots shared in Viber groups, in the "52-week high" figure the trading app itself prominently displays — but it carries no actual information about intrinsic value.

The purchase-price anchor works identically but personally. An investor who bought a commercial bank script at Rs 450 and watches it fall to Rs 310 will very often refuse to sell "until it gets back to at least what I paid," treating Rs 450 as a target the market somehow owes them, rather than evaluating Rs 310 today on its own merits — comparing it, for instance, to the bank's current book value, its dividend trajectory, or better opportunities elsewhere in the market. The market has no memory of what any individual investor paid. It does not know and does not care about your purchase price. Yet the purchase price becomes, in the investor's mind, indistinguishable from "fair value," simply because it is the number most emotionally salient to them.

Anchoring also works in the opposite direction during a rally: an investor who bought a script at Rs 200 and sees it hit Rs 260 anchors to the recent low and concludes the stock is now "expensive," selling far too early relative to the company's genuine growth trajectory, purely because Rs 260 feels large compared to Rs 200 rather than small compared to where sustained earnings growth might eventually take the price.

The practical cost of anchoring is that it substitutes an arbitrary reference point for genuine analysis, in both directions. Anchored to a past high, an investor holds a fundamentally impaired stock far longer than the business justifies, waiting for a "recovery" that may never come because the conditions that produced the old high (a speculative bull run, an industry-wide re-rating that has since reversed, an earnings figure later restated) no longer exist. Anchored to a purchase price, the same investor refuses to realise a loss even when the capital could be redeployed into a demonstrably better opportunity — a subject explored further in Lesson 51.4. The discipline that counters anchoring is deliberately re-deriving a price target from first principles — current earnings, current book value, current growth outlook — and asking "if I did not already own this stock and had never seen its price history, would I buy it today at this price?" If the honest answer is no, the old high or the old purchase price is irrelevant to what you should do next.

Lesson 51.3 — Confirmation Bias and the Telegram Echo Chamber

Confirmation bias is the tendency to search for, interpret, and recall information in a way that confirms what you already believe, while ignoring or discounting information that contradicts it. It is one of the most thoroughly documented biases in all of psychology, and it operates largely outside conscious awareness — investors experiencing confirmation bias genuinely believe they are being objective, because the selective process happens before the information even reaches conscious deliberation.

The mechanism has a simple emotional root: once money is committed to a position, the investor has a psychological stake in that position being correct, separate from the financial stake. Admitting the position was a mistake carries a sting of regret and, often, a public loss of face if the investment was recommended to friends or family or discussed openly in a group chat. To avoid that discomfort, the mind unconsciously filters incoming information — seeking out sources that praise the stock, dismissing sources that criticise it as "haters" or people who "don't understand the sector," and interpreting genuinely ambiguous news (a delayed project completion, a management change, a regulatory circular) in the most favourable possible light.

Nepal's investing culture provides an almost laboratory-perfect environment for confirmation bias to flourish, because so much investment discussion happens inside closed, self-selecting communities: Viber groups organised around a specific broker, a specific sector, or a specific self-styled "guru"; Telegram channels that post buy calls; YouTube channels and Facebook pages whose business model depends on maintaining an enthusiastic, returning audience. These groups are structurally echo chambers, an environment in which the same viewpoint is repeated back to its members so often that it is mistaken for independent confirmation. An investor who has bought a finance company script based on a Telegram "target Rs 900" call will, entirely naturally, gravitate toward the parts of that same channel — and similar channels — that continue to reinforce the bullish thesis, while a sober quarterly report showing rising non-performing loans, published on the company's own website or disclosed to NEPSE, goes unread or is explained away as "temporary" or "the auditors are too conservative." Members who raise doubts in these groups are frequently shouted down or removed, which further purifies the group into a chamber of pure confirmation.

The practical cost of confirmation bias compounds over time in a specific way: it delays the recognition of deteriorating fundamentals, so that by the time the investor finally acknowledges a problem — often forced to by a price collapse too large to explain away — a much larger portion of the loss has already occurred than would have been the case with an early, honest reassessment. Confirmation bias is also the mechanism that makes almost every other bias in this chapter worse, because it prevents corrective feedback from reaching the investor's decision-making in time to matter. A herd-driven purchase, an anchored refusal to sell, an overconfident bet — all of these persist far longer than they otherwise would when the investor is only consuming information that agrees with the original decision.

Lesson 51.4 — Loss Aversion and the Disposition Effect: Why Losers Get Held and Winners Get Sold

Loss aversion is the finding, first rigorously demonstrated by psychologists Daniel Kahneman and Amos Tversky, that losses are felt roughly twice as intensely as equivalent gains are enjoyed. Losing Rs 10,000 produces a sharper, more lasting emotional pain than gaining Rs 10,000 produces pleasure. This asymmetry is not a character flaw; it is a deeply embedded feature of human psychology, likely rooted in an evolutionary history in which the cost of missing a threat (a predator, a food shortage) was far more dangerous than the cost of missing an equivalent opportunity. A useful everyday analogy is spoiled food in the refrigerator: a household will often keep a container of rice or dal that has visibly started to turn, reheating it "just in case," rather than throwing it away — not because anyone genuinely believes eating it is a good idea, but because throwing it away feels like admitting the money spent on it was wasted, and that admission itself is painful, separate from the actual cost of the food. The investor holding a losing stock is doing the identical thing: refusing to "throw away" the position because selling converts an unrealised, abstract loss into a realised, undeniable one, and that psychological finality is what the mind resists — even though the money was already lost the moment the stock price fell, whether or not the position is sold.

This asymmetry produces a specific, well-documented market pattern called the disposition effect: the tendency of investors to sell winning positions too early, to lock in the pleasant feeling of a realised gain, while holding losing positions too long, to avoid the painful feeling of a realised loss. A phenomenological study of Nepali investor experiences has documented this pattern directly among NEPSE participants, describing investors who consistently described selling profitable positions "to be safe" after relatively modest gains, while holding loss-making positions for years, waiting for a breakeven point that in many cases never arrived, particularly among companies whose fundamentals had genuinely and permanently deteriorated.

Picture two investors. The first buys a commercial bank script at Rs 400; it rises to Rs 460, a comfortable 15 percent gain, and the investor sells immediately, satisfied with having "booked profit." The second buys a hydropower script at Rs 400; it falls to Rs 280, a 30 percent loss, and the investor holds on for three years, adding small amounts on the way down "to average the cost," waiting for the price to merely return to Rs 400 so the position can be closed "without a loss." Both decisions are driven by the identical psychological force — the discomfort of realising a loss versus the comfort of realising a gain — but the financial consequences are opposite and severe: the investor systematically prunes their winners while their capital becomes increasingly concentrated in their worst-performing ideas. Over an investing lifetime, this pattern is one of the most reliable destroyers of compounding available, because it guarantees that a portfolio's average holding is disproportionately weighted toward businesses the investor's own actions have already flagged as disappointments.

The single most useful mental correction here is to separate the decision to sell from the price at which you originally bought. Every holding, every single day, should be re-evaluated on one question only: knowing what I know now, would I buy this position today, at today's price, with fresh capital? If the answer is no, the position should be reduced or closed — regardless of whether that crystallises a gain or a loss — because the original purchase price is a sunk cost with zero bearing on the correct decision going forward. This reframing is difficult precisely because it runs directly against loss aversion, which is why it must be built into a written, rules-based process rather than left to be decided fresh, under emotional pressure, each time.

Lesson 51.5 — Overconfidence, Recency Bias, and FOMO: The Bull Market Trio

Three biases tend to arrive together, feed on each other, and do the greatest damage specifically during the euphoric phase of a bull market, which is why this lesson treats them as a connected group rather than in isolation.

Overconfidence bias is the tendency to overestimate the accuracy of one's own knowledge, judgment, and predictive ability. In investing, overconfidence typically arrives on the back of a genuine early success — an investor's first few trades happen to go well, often because they entered during a rising market where nearly everything was rising, and the investor attributes the gain entirely to their own skill in stock selection rather than to the broader tailwind. A study examining risk tolerance, overconfidence, and investment decisions among Nepali investors found overconfidence to be a measurable and significant driver of trading behaviour — investors who scored higher on overconfidence measures traded more frequently, took larger position sizes, and used more leverage (margin lending) than the evidence justified. The mechanism is intuitive: after a lucky win, the brain constructs a flattering narrative — "I have a good instinct for hydropower stocks," "I can time the market" — and that narrative then licenses increasingly large and increasingly under-researched bets, because the investor believes their own judgment alone is sufficient due diligence.

Recency bias is the tendency to give disproportionate weight to recent events and to extrapolate a recent trend forward as though it will continue indefinitely, while discounting longer historical patterns that suggest otherwise. If NEPSE has risen for eight consecutive weeks, recency bias produces the conviction that it will keep rising, because the eight weeks of evidence in front of the investor feel more real and more relevant than the abstract historical knowledge that markets move in cycles and that every sustained rally in NEPSE's history has eventually reversed. Recency bias is what allows an investor to look at a hydropower script's chart, see a clean upward line over the past two months, and conclude the trend is now a reliable fact about the stock's future, rather than one phase of a much longer and more volatile history.

FOMO — fear of missing out — is the anxious, urgent feeling of being left behind while others profit, and it is the emotional accelerant that turns overconfidence and recency bias into actual buying pressure. FOMO is what drives an investor who had originally decided to "wait for a pullback" to abandon that discipline and buy at the top, because every day of waiting brings fresh evidence — friends posting screenshots of gains, a Viber group buzzing with excitement, a script hitting its daily upper circuit for the third day running — that the decision to wait was already a costly mistake. The social dimension matters enormously in Nepal's tightly networked investing culture: hearing that a cousin, a colleague, or a neighbour has made a substantial profit on a specific script in a matter of weeks is a far more emotionally potent trigger than any amount of abstract statistical reasoning about valuation.

Together, these three biases describe the typical arc of a NEPSE retail investor during a bull phase: an early, partly lucky win breeds overconfidence; overconfidence combines with a rising chart to produce recency bias, the belief that the trend will simply continue; and FOMO supplies the emotional urgency to keep buying — and to keep buying larger amounts, often using margin lending to amplify the position — right up until the point where the rally exhausts itself. This is close to the mechanism widely believed to have driven Nepal's dramatic NEPSE cycle beginning in mid-2020, when the index rose from roughly the 1,100 range to an all-time high near 3,200 by August 2021, before falling sharply over the following year as margin calls forced liquidation into a market with progressively fewer willing buyers. Investors who entered late in that cycle — precisely the investors most driven by FOMO, since the loudest social proof and the most dramatic recent gains arrive only after most of the rally has already happened — were disproportionately represented among those left holding positions purchased at or near the peak.

The practical cost of this trio is the most severe in the chapter, because it specifically times capital deployment to the worst possible moment: overconfidence and recency bias encourage progressively larger bets as a rally matures, and FOMO ensures the largest, most leveraged bets are placed closest to the top, right before the reversal that every extended rally in NEPSE's history has eventually produced.

Lesson 51.6 — Gambler's Fallacy and Home Bias: Two Ways of Getting Diversification Wrong

The final two biases in this chapter operate differently from the previous six — they are less about emotional urgency and more about faulty statistical intuition and comfort-seeking — but both quietly erode a portfolio's risk-adjusted return in ways investors rarely notice until the damage is done.

Gambler's fallacy is the mistaken belief that if something has deviated from its average for a period of time, it is now "due" to revert, as though random or semi-random events carry a memory of their own past outcomes. The name comes from the casino: a roulette player who has watched red come up five times in a row often feels black is now overdue, even though the wheel has no memory and each spin remains independent, with the same odds as before. In NEPSE, gambler's fallacy shows up in a very specific, very common piece of reasoning: "this stock has fallen so much, it must bounce back." A script that has fallen from Rs 600 to Rs 150 is treated as statistically "due" for a recovery purely because of the magnitude of the fall, entirely independent of whether anything about the underlying business has stabilised. This is a serious logical error, because a stock price, unlike a roulette wheel, is not a random process oscillating around a fixed mean — it is a reflection of a real business whose value can permanently decline, stagnate, or even go to zero. A finance company that has lost its lending licence, a hydropower project whose river-flow forecasts turn out to have been overstated, or a hotel company that never recovers its pre-pandemic occupancy does not "revert" simply because its stock has already fallen a great deal; a large fall can just as easily be the market correctly repricing a permanently impaired business as it can be an overreaction ripe for reversal, and gambler's fallacy provides no way to tell the two apart.

Home bias, sometimes called familiarity bias, is the tendency to overweight investments in things that feel familiar — domestically, sectorally, or simply by name recognition — relative to what a properly diversified, risk-adjusted portfolio would suggest. Internationally, home bias usually refers to investors overweighting their own country's stock market relative to global markets. In the Nepali context, where capital controls and regulatory restrictions make meaningful international diversification largely inaccessible to ordinary retail investors in any case, home bias manifests one level down: as a heavy overweighting of the handful of sectors and company names an investor already recognises — overwhelmingly, commercial banks, because nearly every Nepali household already has a banking relationship and therefore a comfortable, familiar mental model of what a bank "is" — while sectors that are less intuitively familiar (manufacturing, trading companies, less prominent hydropower developers, insurance) are underexplored or ignored altogether, not because their risk-adjusted prospects are worse, but simply because they are less familiar. A retail investor with a portfolio containing five different commercial bank scripts, and nothing else, often believes they are diversified, because five is more than one. In reality, all five holdings are exposed to the same macro risk factors — interest rate cycles set by Nepal Rastra Bank, the same regulatory capital requirements, the same broad economic cycle, the same NEPSE liquidity conditions — so the portfolio behaves, in a downturn, much more like a single concentrated bet on "Nepali banking" than like a genuinely diversified set of five independent holdings.

The table below summarises all eight biases covered in this chapter, their psychological trigger, and how each typically shows up on NEPSE.

The practical cost of both gambler's fallacy and home bias is quieter than the costs described in earlier lessons, but no less real over time. Gambler's fallacy channels capital into "cheap-looking" but fundamentally impaired businesses on the mistaken belief that a large price fall is itself evidence of an impending recovery. Home bias caps the ceiling on a portfolio's risk-adjusted returns by concentrating exposure in a small number of correlated sectors, leaving the investor far more exposed to a single adverse event — a banking-sector liquidity crunch, a monetary policy tightening cycle — than a properly diversified portfolio spanning multiple, less-correlated sectors would be. Both biases share a common root: they replace an honest assessment of an individual business's prospects with a comforting shortcut — "it must bounce back," "banks are safe because I understand them" — that requires no further work and produces no discomfort, right up until the moment the market disagrees.