The Liquidity-Based Entry and Exit Playbook

First published 24 Aug 2026 · Last verified 29 Aug 2026

The Kaskikot Circuit

Sunita Rana had done everything Part XI told her to do, or so she believed. She had read the chapters on liquidity as a first-order risk, she could recite the ADV rule from memory, and she understood, in the abstract, why free-float allocation limits mattered more than price-earnings ratios for a certain class of NEPSE counters. What she had not done, in October of the previous year, was translate any of that theory into an order she actually typed into her broker's TMS terminal. She had a target: build a position of 60,000 shares in Mustang Resorts Ltd, a thinly traded tourism-and-hospitality counter trading at roughly NPR 480, with a listed base of 40 lakh shares of which barely 12 lakh moved freely outside promoter lock-in. She had a thesis: post-pandemic tourism recovery, a coming dividend announcement, a technical breakout. What she did not have was a plan for how to get in, and, more dangerously, no plan at all for how she would get out.

She bought the first 15,000 shares in a single morning. The order book had perhaps 3,000 shares resting on the ask across the first four price levels. Her market order walked through all of them and kept going, dragging the last price up nearly 9 percent before the session closed. By early afternoon the counter was bid-locked near the upper circuit, other retail accounts piling in behind her, sharesansar forums lighting up with screenshots of the day's gainers list. Sunita felt briefly brilliant. Three weeks later, when tourism arrival numbers came in soft and she needed to raise cash for an unrelated family obligation, she discovered what the other side of that same thinness felt like. Her sell order for even 5,000 shares found no depth at all above the lower circuit. She was locked in a counter she now needed to exit, watching the price gap down two sessions in a row with her order sitting unfilled at the back of an ask queue nobody was buying into.

This chapter is not a retelling of Sunita's mistake for its own sake. It is the operating manual she wishes she had owned before she placed that first order — the practical, mechanical playbook that turns the liquidity concepts from Part XI (ADV thresholds, free-float ceilings, circuit-trap modelling, exit risk as a distinct category from price risk) into a repeatable sequence of decisions: how to size an entry, how to read a depth ladder before committing capital, how to choose between a market order and a limit order inside NEPSE's Trading Management System, and — the discipline that separates professional position sizing from retail impulse buying — how to write the exit plan on the same day as the entry plan, before a single share is bought. Everything in this chapter assumes the reader already knows why liquidity matters. It exists to answer the narrower, harder question: given that it matters, what exactly do you type into the order box on Monday morning?

Lesson 76.1 — Sizing the Entry Against Average Daily Volume

The single most common error among NEPSE retail and even semi-professional investors is treating position size as a function of conviction and capital alone, with average daily volume entering the calculation only as an afterthought, if at all. The ADV rule established in Part XI stated the principle: no single order should represent so large a share of a counter's trading activity that the act of placing it materially moves the price against the person placing it. Lesson 76.1 turns that principle into arithmetic.

Start with the 20-session average daily volume for the counter, not the 250-session average and not yesterday's volume in isolation. A 20-session window is long enough to smooth out a single news-driven spike but short enough to reflect the counter's current liquidity regime rather than a liquidity profile from eight months ago that may no longer hold. For Mustang Resorts, the 20-session ADV that Sunita should have pulled before her first order was approximately 8,000 shares a day — a number available from any standard TMS charting panel or from the historical data section of the exchange's own website, and cross-checked against a merchant banker's research note if one exists for the counter.

The working rule that experienced NEPSE desks apply, and the one this playbook recommends as a default, is that a single day's buy order should not exceed roughly 10 to 15 percent of the 20-session ADV for a counter trading in the illiquid-to-moderate band, rising to perhaps 20 to 25 percent only for counters with genuinely deep, consistently traded floats where the investor has verified depth at multiple price levels on the day itself. Applied to Mustang Resorts, a conservative daily clip is 800 to 1,200 shares; an aggressive one, taken only on days when the book shows unusually deep resting size, tops out near 2,000 shares. Building the full 60,000-share position under the conservative rule requires roughly 50 trading sessions — about ten calendar weeks given NEPSE's five-day trading week. Under the aggressive rule it can be compressed to around 30 sessions, but at a meaningfully higher risk of the investor's own buying becoming a visible, front-runnable pattern to other desk participants watching the tape.

This is where staggering across sessions becomes non-negotiable rather than optional. A single lump-sum order of 60,000 shares in a counter with an 8,000-share ADV is not a large trade in a deep market; it is a market-moving event in a shallow one, and NEPSE's circuit-breaker architecture (discussed fully in Lesson 76.5, but relevant here) means that a determined buyer with real size behind a market order can push a stock through its full daily band and lock it at the upper circuit before the position is even a third built — at which point the remaining two-thirds must be acquired at a price the investor's own buying created, a classic case of an investor becoming their own counterparty risk.

The staggering discipline itself has a simple structure. Divide the target position into weekly tranches rather than daily fixed quantities, because NEPSE liquidity is not evenly distributed across the week — Sunday and Monday sessions (the NEPSE week runs Sunday through Thursday) frequently carry different volume characteristics than Wednesday and Thursday sessions, often thinner at the week's open as participants recalibrate after the weekend gap, thickening toward midweek. Within each weekly tranche, place orders only on sessions where the pre-open indicative price and the first fifteen minutes of trade show the counter is not already gapping toward its circuit band; skip a session entirely rather than force an entry into a already-moving tape. This single habit — the willingness to simply not trade on a given day — is the most underused tool in the retail NEPSE playbook, precisely because it produces no visible action and therefore feels like inactivity rather than discipline.

A secondary sizing check, drawn directly from the free-float allocation limits covered in Part XI, applies at the position level rather than the daily-order level: what fraction of the counter's total free float will this position represent once fully built? For Mustang Resorts, a 60,000-share position against a free float of roughly 12 lakh shares is 5 percent of the tradable float — a level at which Sunita is not yet a market-moving holder in the regulatory sense, but is large enough that her own eventual exit will itself be a liquidity event the size of many single trading sessions combined. That fact should be visible to her on day one, not discovered on the day she needs to sell.

Table 76.1 sets out an operating grid connecting ADV, target position size, and the corresponding entry runway under the conservative 12 percent daily-clip assumption. It is meant as a working reference, not a rigid formula — the constants should be adjusted for the investor's own risk tolerance and for the specific counter's depth behaviour.

Sunita's 60,000-share target against an 8,000-share ADV placed her at 7.5x ADV — squarely in the moderate band, entry-feasible but demanding real patience, and, as she discovered, demanding an exit plan of at least equal seriousness to the entry plan. Had her target been 120,000 shares, the grid would have told her before she placed a single order that the position sat in the severe band and should have been capped or abandoned at the outset.

Lesson 76.2 — Reading the Depth Ladder Before You Click

Every order placed on NEPSE through a broker's TMS terminal interacts with a visible order book: a ladder of resting buy orders (bids) below the last traded price and resting sell orders (asks) above it, typically displayed five levels deep on most retail TMS interfaces, sometimes more on desktop or professional terminals. Reading that ladder correctly, before placing an order rather than after watching the fill, is the second load-bearing habit in this playbook, and it is the one Sunita skipped entirely on her first Mustang Resorts order.

The depth ladder answers three questions that price alone cannot answer. First, how much size actually rests at each level, on both the bid and the ask side — a stock can show a seemingly reasonable last-traded price while carrying almost no resting size behind it, meaning the next order of any real quantity will move the price several ticks regardless of the trader's intention. Second, how the size is distributed — a ladder with 200 shares at the best ask and 4,000 shares bunched three levels up behaves very differently from one with 4,000 shares spread evenly across all five levels; the former will let a small order through cleanly but will gap violently on anything larger, while the latter absorbs size more gradually. Third, how the ladder is behaving dynamically over the pre-order minutes — orders appearing and disappearing at the top of book (a pattern sometimes associated with orders being placed and pulled to test reaction, colloquially referred to by NEPSE retail traders as book manipulation, though it is difficult to prove and this playbook does not ask the reader to diagnose intent, only to notice the pattern and treat it as a reason for caution) is a different signal than a stable ladder that holds its shape.

The mechanical rule this playbook recommends: before placing any order larger than roughly 5 percent of a session's likely volume, open the depth ladder and estimate what economists would call the order's expected slippage — the gap between the best available price and the average price the full order would achieve if it consumed every level of resting size in its path. If that estimated slippage exceeds roughly 1.5 to 2 percent of the current price, the order is too large for the moment's depth and should be split further or deferred to a session with thicker resting size, irrespective of what the daily-ADV arithmetic from Lesson 76.1 already suggested.

This is also the discipline that prevents the single most damaging entry mistake in a low-float counter: chasing the circuit. NEPSE's daily price band — currently set at 15 percent for an individual scrip's movement in a session, following the exchange's most recent circuit-breaker revision, with a market-wide trading halt triggered if the benchmark index itself swings roughly 8 percent intraday — creates a specific, recurring trap for retail entrants. When a thinly traded counter begins moving toward its upper circuit on real or rumoured news, the visible ask-side depth thins rapidly as existing holders pull offers rather than sell into a rising market, while buy-side orders pile up faster than the remaining size can absorb them. An investor who places a market order into that environment is not buying at the last traded price; they are buying at whatever price the last remaining seller in the queue was willing to accept, which in a circuit-locked NEPSE counter frequently means the order sits unfilled at the top of a buy queue behind dozens of other buyers, executes only a token quantity through the day's closing call-auction mechanism, and leaves the investor holding an unfilled order rolling into the next session — at a price that has now already moved 15 percent from where the thesis was formed.

The rule that follows directly from this pattern: never initiate or add to a position in a counter that is already within one or two ticks of its upper circuit for the session, no matter how strong the underlying conviction. The correct response to a stock gapping toward its circuit is to wait — either for the circuit to lock and the session to end, giving a full day's distance to assess whether the move reflects durable news or a liquidity-driven spike, or for the ladder to show genuine two-sided depth returning rather than one-sided buy pressure against vanishing offers. Capital not deployed today into a chasing trade is capital still available tomorrow at a readable price; capital deployed into a circuit chase is capital whose entry price was set by someone else's urgency, not the investor's own analysis.

Lesson 76.3 — Writing the Exit Plan Before the First Share Is Bought

If Lessons 76.1 and 76.2 govern the mechanics of getting into a position without damaging the entry price, Lesson 76.3 addresses the discipline that most reliably separates investors who survive a NEPSE liquidity event from those who do not: the exit plan must be written, in specific and numeric terms, before the entry order is placed — not after the position is built, and certainly not after the position has already moved against the investor and emotion has entered the decision.

The reasoning is straightforward once stated, though it runs against a strong behavioural current. An exit plan written before entry is written by an investor with no position-specific emotional stake yet formed — no anchoring to a purchase price, no sunk-cost pull toward holding a loser to "get back to even," no euphoria-driven temptation to let a winner run past its rational target because the position feels validated. An exit plan written after the position already exists and has moved is written by a different psychological actor entirely, one whose analysis of the correct exit price is contaminated by where the price currently sits relative to the entry, which is information the market does not care about and should not be part of a rational sell decision.

The practical form this takes for a NEPSE position: at the same session the entry plan is drafted (as in Lesson 76.1, before the first tranche is bought), the investor writes down a scaled exit schedule keyed to price levels, not to time, and separately writes down a liquidity-based exit schedule keyed to session counts, not to price. The two schedules serve different purposes and must both exist.

The price-keyed schedule answers "at what valuation do I reduce or close this position." For Sunita's Mustang Resorts thesis, a disciplined version drafted at entry might have read: sell 25 percent of the position at a 20 percent gain from average cost, sell a further 35 percent at a 40 percent gain, and hold the remaining 40 percent against a trailing stop set at 15 percent below the highest close achieved once the position has cleared its first profit tranche. None of these numbers are magic; the specific percentages should reflect the investor's own risk tolerance and the thesis's own return expectations. What matters is that the numbers exist in writing before entry, so that when the price actually reaches NPR 576 (a 20 percent gain from a 480 average cost) some weeks later, the decision to sell a quarter of the position is a mechanical execution of a pre-committed plan rather than a fresh, emotionally loaded decision made in the moment with a green position on the screen creating its own psychological pull to let it run further.

The liquidity-keyed schedule answers a different, and for illiquid NEPSE counters, more urgent question: "given this counter's ADV, how many sessions will a full exit actually require, and does my capital-need timeline allow for that." This is the schedule Sunita never wrote, and its absence is what turned a manageable position into a trapped one. The same ADV arithmetic from Lesson 76.1 applies in reverse: if a full exit of 60,000 shares, capped at 12 to 15 percent of a 20-session ADV of 8,000 shares per session, requires a comparable 40-to-50-session runway, then any capital need the investor anticipates on a shorter horizon than that is a signal that the position is oversized relative to the investor's own liquidity requirements, not merely relative to the stock's trading characteristics. This is a position-sizing conclusion, and it needs to be reached before entry, when the position size is still an adjustable variable, rather than after entry, when it has become a fixed constraint the investor is now negotiating against under duress.

The exit plan should also specify, in advance, what triggers an acceleration of the schedule beyond the routine price-and-liquidity framework — a genuine deterioration in the investment thesis (a weak tourism season, a governance concern, a sector-wide regulatory shift of the kind covered in earlier chapters on manufacturing and hospitality counters) as distinct from ordinary price volatility that does not change the underlying case. Writing this distinction down at entry prevents the common failure mode of an investor deciding, mid-decline, that the thesis has changed simply because the price has, when in fact nothing about the business has moved and the price action itself is the liquidity-driven noise this entire playbook is designed to look past.

Lesson 76.4 — Unwinding a Large Position in a Low-Float Counter Without Crashing It

Lesson 76.3 established that the exit plan exists in writing before entry. Lesson 76.4 addresses the harder practical case: what an investor actually does, mechanically, when the exit plan is triggered and the position is large relative to the counter's float and ADV — precisely the situation Sunita faced when she needed to unwind her Mustang Resorts holding under real time pressure.

The core hazard in exiting a large position in a thin counter is the mirror image of the entry hazard from Lesson 76.1, but with an added asymmetry that makes it more dangerous: sellers in a panic or under a deadline tend to reach for market orders far more readily than buyers do, because the psychological pressure of needing cash by a date feels more urgent than the pressure of wanting to deploy cash by a date. A market sell order for any meaningful quantity in a counter with a shallow bid ladder will walk down through every resting bid in its path, and because NEPSE's circuit architecture applies symmetrically, a large enough sell order can drive a counter to its lower circuit within a single session, at which point — exactly as happened to Sunita — the remaining unsold shares are stuck behind a wall of an already-locked price with no fresh bids appearing, and the investor's own selling pressure has created the very illiquidity trap that then prevents further selling.

The correct sequence, once an exit is triggered, mirrors the staggered-entry discipline from Lesson 76.1 but adds several exit-specific refinements.

First, the exit should begin from the top of the pre-written scaled schedule (Lesson 76.3) rather than being dumped in full immediately upon the trigger, even under real time pressure, because a full-size market exit is precisely the action most likely to produce the worst possible average price. An investor who needs cash in three weeks is generally better served by an aggressive but still staggered exit over the first ten to fifteen sessions of that window than by a single-day liquidation, even though the single-day liquidation feels psychologically like it resolves the uncertainty faster.

Second, exit orders in a thin counter should be placed as limit orders resting slightly above the current best bid — not as market orders, and not as limit orders priced so far above the market that they never fill. The specific placement (discussed further in Lesson 76.5) allows the investor to capture whatever genuine buying interest arrives during the session without accepting whatever price a market order would produce by consuming the full visible bid ladder in one pass.

Third, and this is the refinement most retail investors miss entirely, the exit should be paced to be less conspicuous than the daily-ADV ceiling alone would suggest, because a seller placing the maximum allowable size at the same time each session, day after day, becomes a recognizable pattern to other market participants watching the tape — brokers' proprietary desks, active day traders, and algorithmic-adjacent retail flow all learn to identify a large, patient, mechanical seller and adjust their own bidding downward in anticipation, a phenomenon sometimes called information leakage in market-microstructure terms. Varying the size and timing of exit tranches within the ADV ceiling — some sessions at 8 percent of ADV, others at 13 percent, placed at different points in the trading window rather than always at the open — reduces the degree to which the investor's own selling becomes predictable and therefore exploitable.

Fourth, and specific to the circuit-trap risk this playbook inherits directly from Part XI's exit-risk modelling, the investor must monitor whether their own selling, combined with any independent negative sentiment, is pushing the counter toward its lower circuit band. If a session opens with the stock already gapping down and thin bid-side depth, the correct response — counterintuitively for an investor under time pressure — is often to skip that session's planned sell tranche entirely rather than force it, because selling into a circuit-bound decline both achieves a worse price than waiting and increases the probability of contributing to a lock that then prevents any further selling that day or the next. The liquidity-runway math from Lesson 76.3 should have already built in a buffer of several extra sessions precisely to absorb one or two skipped days without breaching the capital-need deadline; an exit plan with zero slack for a bad session is an exit plan that was underbuilt from the start.

Table 76.2 summarises the exit-side decision framework across the range of scenarios an investor is likely to face, distinguishing routine profit-taking from deadline-driven and deterioration-driven exits.

Sunita's actual unwind, once she applied this framework retroactively with a broker's guidance, took nineteen sessions rather than the single catastrophic day she had initially attempted, and achieved an average exit price roughly 9 percent better than the price she would have realised had she continued forcing market sells into a thin, deteriorating bid ladder. It also required her to renegotiate her family's cash-need timeline by roughly three weeks — an outcome only possible because she communicated the liquidity constraint honestly and early, rather than discovering it under maximum pressure with no time left to adjust either the exit pace or the deadline.

Lesson 76.5 — Market Orders and Limit Orders Inside NEPSE's TMS

Every rule established so far in this chapter ultimately resolves into a single mechanical choice made at the point of order entry inside NEPSE's Trading Management System: whether to submit a market order or a limit order, and, within a limit order, how to set its validity and its price relative to the current book. This lesson makes that choice explicit.

NEPSE trading is conducted electronically through TMS, the exchange's order-matching and trading infrastructure, accessed by retail investors through broker-provided web and mobile front ends rather than a direct exchange terminal. Every order placed through these front ends carries two independent attributes: an order type, governing how the price is determined, and a validity condition, governing how long the order remains live if it does not fill immediately.

The order type choice, in NEPSE's TMS terminology, is typically presented to the investor as MKT (market) or LMT (limit). A market order instructs the system to execute immediately against the best available opposing price on the book, consuming as many price levels of depth as necessary to complete the requested quantity, with no ceiling on how far the execution price can move from the last traded price short of the day's circuit band itself. A limit order instructs the system to execute only at the investor's specified price or better, resting on the book unfilled if no matching opposing order exists at that price, and never executing at a worse price than specified regardless of how thin the book is.

For the overwhelming majority of situations this chapter addresses — staggered entries into a low-ADV counter, and especially scaled exits from a large position in a thin counter — the limit order is the structurally correct default, and the market order should be treated as an exception reserved for genuinely liquid, deep-book counters where the investor has verified that even a market order's worst-case slippage is trivial. The reason follows directly from everything established in Lessons 76.1 through 76.4: a market order in an illiquid counter hands full control of the execution price to whatever happens to be resting on the opposing side of the book at that instant, which is precisely the outcome this playbook exists to prevent.

Within the limit-order choice, placement relative to the current book matters as much as the type itself. A limit buy placed exactly at the best current ask will fill immediately if size exists there, functioning close to a market order in effect but with a hard price ceiling; a limit buy placed one or two ticks below the best ask waits for the market to come to it, filling only if a seller is willing to meet that price, which costs execution certainty but frequently achieves a materially better average price across a multi-session campaign. For entries under this playbook's staggered framework, placing limit buys at or just inside the current best bid — rather than lifting the ask — is generally the more disciplined approach, accepting that some sessions may not fill at all rather than forcing a fill by paying up.

This variability matters directly for the staggered-entry and scaled-exit disciplines built up across this chapter: an investor whose broker platform supports only day orders must manually re-enter each tranche's limit order every single session, since anything unfilled simply expires at close rather than carrying forward — a real operational burden, but one that is far preferable to defaulting to market orders purely to avoid the inconvenience of re-entering a limit order each morning. Where a multi-session validity type is available, it can reduce that operational burden, but the investor should still review and, where needed, revise the resting price each session rather than leaving a stale limit order unattended for days at a time in a moving market.

Lesson 76.6 — The Entry and Exit Decision Tree

The five lessons preceding this one each address one component of the liquidity-based playbook in isolation: sizing against ADV, reading depth, pre-writing the exit, unwinding without a crash, and choosing the right order mechanics inside TMS. Lesson 76.6 assembles these into a single decision sequence an investor can actually run, in order, every time a position is contemplated, entered, held, or exited.

Before any capital is committed, the sequence begins with data gathering, not with an order screen. The investor pulls the counter's 20-session ADV, checks it against the target position size to classify the position on the runway grid from Table 76.1, and checks the target position size against the counter's estimated free float to flag whether the position sits at a level that will itself constitute a meaningful liquidity event on exit. If the position classifies as severe on that grid — above roughly 15x ADV — the correct action is to reduce the target size or decline the position before any order is placed, not to proceed and hope the arithmetic proves pessimistic.

If the position clears that first gate, the investor next drafts the exit plan in full: the price-keyed scaled schedule, the liquidity-keyed session-count runway including the T+2 settlement buffer, and the acceleration triggers that distinguish thesis deterioration from ordinary volatility. This plan is written and, ideally, recorded somewhere the investor will actually revisit — a dated note, a simple spreadsheet row, a line in a trading journal — precisely so that it exists as a pre-commitment device rather than a memory that conveniently reshapes itself once the position has moved.

Only then does the entry sequence itself begin. Each session, before placing an order, the investor checks the depth ladder for genuine two-sided size and estimates slippage for the day's planned tranche; if the counter is already gapping toward its upper circuit or the ladder shows one-sided, thinning depth, the session's tranche is skipped rather than forced. Where the ladder supports it, the tranche is placed as a limit order priced at or just inside the current bid, sized within the ADV ceiling established in Lesson 76.1, and the cumulative position is tracked against the target so the investor always knows what fraction of the plan remains.

Once the position is fully built, the sequence shifts to monitoring against the pre-written exit triggers rather than to any new decision-making about whether to hold or sell — that decision was already made, in writing, before entry. When a price target from the scaled schedule is reached, or a liquidity-need deadline approaches, or a genuine thesis-deterioration trigger fires, the exit sequence from Lesson 76.4 executes: staggered tranches within the ADV ceiling, limit orders rather than market orders, session-skipping when the bid ladder thins or the counter approaches its lower circuit, and continuous comparison of the remaining runway against whatever deadline is driving the exit.

The decision tree resolves, at every stage, to one of a small number of concrete actions: proceed with today's planned tranche, reduce today's tranche size, skip today's session entirely, or — at the position-sizing gate before entry — decline the position altogether. None of these actions require sophisticated tools beyond a broker's standard TMS interface and a willingness to check the depth ladder and the ADV figures before, not after, placing an order. The discipline this chapter asks for is almost entirely procedural rather than analytical: the analysis of whether Mustang Resorts or any other counter is a good investment belongs to other chapters in this book. This chapter's entire claim is narrower and, in practice, more consequential for capital preservation — that even a correct thesis, badly sized and badly sequenced against a thin counter's liquidity, can produce a worse outcome than a mediocre thesis executed with genuine liquidity discipline.