The Philosophy of Valuation

First published 23 Aug 2026 · Last verified 29 Aug 2026

Lesson 45.1 — Price Is What You See; Value Is What You Are Trying to Find

Walk through the Ason vegetable market in Kathmandu on any given morning and you will see the same tomato selling for three different prices at three different stalls, all within fifty meters of each other. A tourist pays the highest price. A regular customer who knows the vendor pays the middle price. A wholesaler buying forty kilograms for a hotel pays the lowest price per kilogram. The tomato itself has not changed. Its nutritional content, its ripeness, its usefulness in a meal — its "worth" to whoever eats it — is identical across all three transactions. What changed was the price, and price moved for reasons that had nothing to do with the tomato's underlying worth: bargaining skill, information, urgency, relationship, and quantity.

This distinction — between what something costs and what it is actually worth — is the single most important idea in this Part of the book, and arguably in all of investing. It is often summarised in a line popularly attributed to Warren Buffett, paraphrasing his teacher Benjamin Graham: "Price is what you pay; value is what you get." The idea did not originate with Buffett — Graham's 1949 book *The Intelligent Investor* built an entire philosophy around the gap between a security's market price and its underlying worth — but the one-line version has become the most quoted sentence in investing precisely because it captures something every investor eventually learns the hard way.

Apply this to a share of stock instead of a tomato. The price of a share of Nabil Bank or Chilime Hydropower or Unilever Nepal on the NEPSE (Nepal Stock Exchange) trading screen at 11:45 AM on a Tuesday is simply the level at which the most recent buyer and seller agreed to transact. It tells you what someone was willing to pay a moment ago. It does not, by itself, tell you what the underlying business — its factories, its licenses, its deposit base, its power purchase agreements, its brand, its future stream of profits — is actually worth to a rational, well-informed owner who intends to hold it for years. Value, in this sense, is an estimate of a business's true underlying worth, built from its assets, earnings power, growth prospects, and risk — not from what the ticker tape says right now.

Why would price and value ever diverge? Because a market price is set by whoever happens to be trading at that moment, and their motives may have nothing to do with a careful assessment of worth. A person needing cash for a medical emergency will sell shares below what they believe those shares are worth, simply because they need money now. A person who just read an exciting rumour on a Viber group will buy shares above what a calm analysis would support, simply because they are excited. A large institution rebalancing its portfolio at month-end may sell a stock not because the business has deteriorated, but because a target allocation percentage says to sell. None of these transactions is "wrong" — they reflect real motives — but none of them is a statement about intrinsic value either.

Intrinsic value is the term this book will use throughout Part IX for that underlying, patiently-estimated worth of a business — the price a rational, informed buyer with a long time horizon should be willing to pay, based on the cash the business can be expected to generate over its life, discounted back to what that future cash is worth today. It is called "intrinsic" precisely to distinguish it from market price: it belongs to the business, not to the crowd's mood about the business on a given day.

This is not an abstract, academic distinction for Nepal. It is arguably more relevant here than in almost any other market an investor is likely to encounter, for reasons this chapter will build toward across its six lessons. But the starting point is simple and must be internalized before anything else in this Part makes sense: your job as a serious investor is never to predict what the price will do tomorrow. It is to estimate, as carefully as the available information allows, what the business is worth — and then to compare that estimate to the price the market is offering. Everything else in valuation is a set of tools for doing that comparison well.

Lesson 45.2 — Why Different Methods Legitimately Disagree

A newcomer to valuation is often unsettled to discover that three competent analysts, looking at the same company on the same day with the same publicly available financial statements, can produce three meaningfully different value estimates — and that this is normal, expected, and not a sign that valuation is fraudulent or useless. Understanding why this happens is essential before you learn any specific method, because without it you will either (a) trust a single number far more than it deserves, or (b) conclude that valuation is a waste of time because "the answer is never the same twice."

Return to the house-appraisal analogy, because it will recur throughout this Part and it maps unusually well onto company valuation. Suppose you want to know what a house in Lalitpur is worth. There are at least three legitimate ways to answer this question, and they do not need to agree:

First, you could look at what similar houses in the same neighbourhood recently sold for — three bedrooms, similar plot size, similar road access — and adjust for differences. This is comparable sales, and in company valuation its equivalent is relative valuation (also called "comps" or multiples-based valuation): looking at what the market is currently paying for similar companies, expressed as a ratio like price-to-earnings or price-to-book, and applying that ratio to the company you are studying.

Second, you could estimate what it would cost, today, to buy the land and rebuild an identical house from scratch — bricks, labor, permits, everything — and call that the value. This is replacement cost or asset-based valuation, and its company-valuation equivalent is book value or net asset value (NAV): adding up what the company's assets are worth (adjusted to reflect reality, not just historical accounting cost) and subtracting what it owes.

Third, if the house is rented out, you could estimate the rental income it will generate over the coming years and calculate what that stream of future rupees is worth in today's money. This is income capitalisation, and its company-valuation equivalent is the discounted cash flow (DCF) method, or, for a bank or financial institution where the relevant "income" to the shareholder is the dividend rather than the whole firm's cash flow, the dividend discount model (DDM).

Three honest, competent appraisers using these three approaches on the same house can and do arrive at three different figures, and none of them is "lying." The comparable-sales appraiser might land higher if the neighbourhood is in a speculative upswing and recent sale prices reflect optimism rather than durable worth. The replacement-cost appraiser might land lower if land prices have risen faster than construction costs, making the existing structure cheap to replicate. The income appraiser might land lowest of all if rental yields in the area are compressed relative to purchase prices. All three numbers can be simultaneously "correct" in the sense that each was calculated correctly given its method and its assumptions — while still disagreeing, because they are measuring different things and relying on different assumptions about the future.

The same is true of company valuation, for the same underlying reason: every method requires assumptions, and reasonable people can disagree about assumptions without either party being incompetent or dishonest. A DCF valuation of a hydropower company depends heavily on the discount rate used (how much return investors demand for the risk involved) and on assumptions about hydrology, tariff escalation, and the terminal value once the power purchase agreement (PPA) period ends. A relative valuation of the same company depends on which "comparable" companies you choose and whether the market's current pricing of those comparables is itself sensible or inflated by a sector-wide enthusiasm. A book-value approach depends on how faithfully the balance sheet reflects economic reality — whether a dam built fifteen years ago at historical cost is actually worth what the accounting entry says today.

This is why Part IX of this book is structured the way it is. This chapter is the philosophical foundation; the chapters that follow will each take up one method in depth — DCF, relative valuation, the dividend discount model, and book value/NAV approaches — precisely because no single tool is sufficient. A carpenter does not show up to a job with only a hammer. A serious investor does not show up to a valuation question with only one method.

Lesson 45.3 — Margin of Safety: Valuation's Purpose Is Not Precision, It Is Protection

Once you accept that any single valuation estimate carries real uncertainty — that your DCF could be too optimistic, your chosen "comparable" companies could themselves be mispriced, your book-value adjustments could be wrong — a natural and important question follows: if valuation is inherently imprecise, why bother at all? The answer, developed most fully by Benjamin Graham and carried forward by generations of practitioners since, is the concept of margin of safety.

Margin of safety means buying a security at a meaningful discount to your estimate of its intrinsic value, so that even if your estimate turns out to be somewhat wrong — even if the business performs worse than expected, or your assumptions prove too rosy — you are still likely to avoid a serious permanent loss, and still have room to profit. It is not a formula; it is a discipline of humility built into the buying decision itself.

Consider a bridge engineer. If an engineer calculates that a bridge needs to hold trucks weighing up to 10 tons, they do not design the bridge to hold exactly 10 tons. They design it to hold 30 tons or more, because they know their models of stress, material fatigue, and load distribution are imperfect, and because the cost of being wrong (a collapsed bridge) is catastrophic and asymmetric compared to the cost of over-building (some excess steel). Margin of safety in investing works the same way: you deliberately demand a price well below your estimated value, not because you doubt your own arithmetic, but because you know every valuation model is a simplification of a more complicated reality, and you want room for that reality to disappoint you without destroying your capital.

Why does this matter more, not less, in Nepal specifically? Because several structural features of the NEPSE make both price volatility and valuation-model error more likely than in a deeply liquid, institutionally-dominated market. Retail investors — individuals trading their own savings rather than professional fund managers — make up the overwhelming majority of NEPSE's trading activity and ownership, a pattern documented repeatedly in research on Nepal's capital market and discussed in earlier chapters of this book covering NEPSE's market structure. A market dominated by individual retail participants, many of them trading part-time alongside other jobs, tends to be more sentiment-driven: prices can swing on rumour, on a single piece of news amplified across social media and investment Viber/Telegram groups, or on simple herd behaviour, rather than on a considered reassessment of a company's earnings power.

Add to this the fact that many NEPSE-listed counters — especially smaller hydropower companies, smaller finance companies, and thinly-traded manufacturing names — suffer from illiquidity: on many trading days, only a small number of shares change hands, so a single moderately-sized buy or sell order can move the price by several percentage points without reflecting any real change in the business. In a liquid market like a major global exchange, heavy two-sided trading by large, well-resourced institutions tends to pull price back toward consensus estimates of value fairly quickly, because mispricing attracts arbitrage capital. NEPSE has far less of this corrective mechanism. A share can trade meaningfully above or below a careful analyst's estimate of intrinsic value for a long time, because there may not be enough patient, analytically-driven capital in the market to close the gap quickly.

The behavioural biases covered in earlier chapters of this book — anchoring on a stock's previous high price, herd behaviour around hot sectors, overconfidence after a run of lucky gains, loss aversion that keeps investors holding falling stocks too long — are not separate from valuation discipline; they are the reason valuation discipline is necessary. A valuation framework is, in a real sense, a structured defence against your own psychology and against the crowd's psychology. When everyone around you is buying a hydropower stock because "it only goes up," a calm DCF estimate of the company's actual cash-generating capacity over the life of its power purchase agreement is one of the few tools that can interrupt that momentum with a genuine, numbers-based question: at this price, am I actually being compensated for the risk I am taking, or am I simply paying for other people's enthusiasm?

Lesson 45.4 — Four Lenses, One House: An Overview of the Methods to Come

Having established why price and value differ, why methods legitimately disagree, and why margin of safety matters especially in a market like NEPSE, this lesson previews the four valuation approaches that the remaining chapters of Part IX will each treat in full depth. Think of this as being handed a map before the detailed hike — enough to see how the pieces relate to one another before you descend into any one of them.

Discounted cash flow (DCF). This method estimates the cash a business is expected to generate for its owners in future years, then discounts those future cash flows back to a present value using a rate that reflects the riskiness of receiving them (the further out and the riskier the cash flow, the more it is discounted). It is the most theoretically complete method, because in principle it captures everything that matters — growth, profitability, capital needs, and risk — in a single framework tied directly to cash the business actually produces. Its weakness is sensitivity: small changes in the assumed growth rate or discount rate can swing the resulting value substantially, so the quality of a DCF is only as good as the quality (and humility) of its inputs.

Relative valuation (comps). This method sidesteps forecasting the distant future and instead asks what the market is currently paying for similar businesses, expressed as a multiple of some financial metric — price-to-earnings (P/E), price-to-book (P/B), EV/EBITDA, and similar ratios — and applies a comparable multiple to the company being valued. Its strength is that it is grounded in real, observable market prices rather than a long chain of speculative assumptions about the future. Its weakness is that it inherits whatever mispricing already exists in the market: if an entire sector on NEPSE is trading at an inflated multiple because of retail enthusiasm, a comps-based valuation of a single company in that sector will simply reproduce that same inflated valuation, dressed up in the language of "the market says."

Dividend discount model (DDM). A specialised cousin of DCF, the DDM values a share based specifically on the stream of dividends a shareholder actually expects to receive, discounted back to present value. It is particularly suited to businesses — commercial banks being the leading NEPSE example — where regulatory capital requirements set by Nepal Rastra Bank (NRB) constrain how much cash can actually leave the business and reach shareholders, meaning dividends (cash and, in Nepal's market, frequently bonus shares) are a more reliable proxy for shareholder value than the bank's broader free cash flow, which is difficult to define cleanly for a financial institution whose "raw material" is deposits and whose "product" is loans.

Book value / net asset value (NAV). This method values a company based on what its assets are worth (adjusted where possible to current, realistic values rather than historical accounting cost) minus its liabilities. It is the natural lens for asset-heavy or asset-backed businesses — banks and financial institutions again (where regulatory capital is explicitly measured against book equity), insurance companies, and to a degree hydropower companies with tangible, licensed infrastructure — and it is also the natural floor-level check for any company in liquidation or distress, where the question shifts from "what can this business earn?" to "what could its assets fetch if sold off piece by piece?"

None of these methods is superior in the abstract. Each is more or less appropriate depending on two things this chapter now turns to: the type of company being valued, and the purpose for which the valuation is being done.

Lesson 45.5 — Matching the Method to the Company

A recurring error among newer analysts is applying a favourite valuation method to every company regardless of fit — running a DCF on a bank the same way one would on a manufacturer, or valuing a hydropower company on price-to-earnings the way one would value a trading company. The method must fit the economics of the business, in the same way an appraiser would not value a working farm the same way as a downtown apartment — the income sources, the assets, and the risks are simply too different.

Consider three archetypal NEPSE-listed company types and why they call for different primary lenses.

A commercial bank. A bank's balance sheet is its business — deposits are its raw material, loans and investments are its output, and its capital adequacy ratio (a regulatory measure of how much shore-up capital it holds against its risk-weighted assets, set and enforced by Nepal Rastra Bank) directly constrains how fast it can grow and how much it can pay out. Trying to build a from-scratch free-cash-flow DCF for a bank is notoriously difficult, because "capital expenditure" and "working capital" do not mean the same thing for a lender that they mean for a factory — a bank's core activity of taking in deposits and making loans blurs the normal DCF distinction between operating cash flow and investing cash flow. For this reason, banks worldwide are conventionally valued primarily through price-to-book (comparing market price to the bank's book equity, adjusted for asset quality) and through the dividend discount model (since a bank's ability to pay dividends is directly gated by its NRB-mandated capital buffers, and Nepali investors have historically prized the bonus shares and cash dividends banks distribute). Relative valuation against other listed commercial banks is also highly informative here, because Nepal has more than two dozen listed banks and finance companies with genuinely comparable business models, regulatory regimes, and disclosure formats, making peer comparison unusually reliable for this sector specifically.

A hydropower company. A hydropower project is close to the opposite case: it is a long-duration, asset-heavy business with a specific, contractually defined life — most Nepali hydropower companies sell electricity to the Nepal Electricity Authority (NEA) under a power purchase agreement (PPA) with a fixed tariff schedule and duration, after which the terms may change or the license may need renewal. This structure — a known (if long) horizon, contractually bounded revenue, and heavy upfront capital investment already sunk into dams, tunnels, and powerhouses — makes hydropower a textbook DCF candidate, much like valuing a toll road or an annuity: forecast the electricity generated (which depends on river flow/hydrology and plant capacity), apply the PPA tariff schedule (with escalation and wet/dry season rate differences common in Nepali PPAs), subtract operating costs and debt service, and discount the resulting cash flows at a rate reflecting hydrology risk, counterparty risk (NEA's own payment reliability), and Nepal's country-level risk. Book value is a much weaker lens here, because a dam's historical construction cost bears little relationship to its ongoing earning power, which depends on rainfall patterns and tariff terms, not on what steel and cement cost when it was built.

A manufacturing or consumer company. A company producing noodles, cement, steel, or beverages for the Nepali market sits closer to the classic case most global valuation textbooks are written for: it has a normal cycle of buying raw materials, converting them into products, selling them, and reinvesting profits into more capacity — with capital expenditure and working capital behaving in the textbook way. This makes it well suited to a standard free-cash-flow DCF and, because Nepal has multiple listed manufacturers and consumer companies of varying scale (some direct sector peers, some approximate), relative valuation using P/E or EV/EBITDA multiples against domestic peers, adjusted where necessary against regional peers when domestic comparables are too thin.

Below is a summary table intended as a working reference, not a rigid rulebook — real companies often benefit from two or three methods used together, as Lesson 45.2 already argued.

Lesson 45.6 — Matching the Method to the Purpose

The type of company is only half the equation. The other half — often neglected by newer analysts — is the purpose of the valuation: what decision is this number actually meant to inform? Valuing a company to decide whether to buy a small minority stake through the NEPSE trading screen is a genuinely different exercise from valuing a company to decide whether to acquire the whole thing, and conflating the two leads to real errors.

When you buy shares of, say, Himalayan Bank or Chilime through your broker's trading terminal, you are buying a minority stake: a small slice of a business you will not control, whose board you cannot appoint, whose dividend policy you cannot set, and whose strategic decisions you can only vote on alongside thousands of other shareholders. For this purpose, the relevant valuation question is essentially: given what a passive minority owner can expect to receive (dividends, bonus shares, and eventually a sale at some future price), and given the price the market is currently asking, is this an attractive risk-adjusted proposition? Relative valuation against listed peers, and dividend-focused approaches, are usually highly relevant here, because minority shareholders' returns are substantially shaped by what the company chooses to distribute and by what the broader market is willing to pay for similar minority stakes — not by any control premium.

Contrast this with valuing a company for a full acquisition, a merger, or a controlling private-equity-style investment — a purpose more relevant to Nepal's banking sector consolidation waves (where NRB has periodically pushed mergers among BFIs to strengthen capital bases) or to family-owned manufacturing and hydropower businesses considering a strategic sale. Here, the buyer is not a passive recipient of whatever dividend policy management chooses; the buyer can change the dividend policy, replace management, alter the capital structure, sell off unproductive assets, or redirect the entire cash flow of the business. This is why acquisition valuations typically rely more heavily on full enterprise DCF (valuing the whole stream of cash the business can generate under new, presumably improved, control) and on NAV/asset-based approaches (because a controlling buyer can actually realise asset value directly — sell a building, monetize a license — in a way a minority shareholder cannot). Acquisition valuations also typically include a control premium: an amount paid above the "minority stake" market price, precisely because control over cash flows and assets is worth more than a passive claim on whatever a company's existing board decides to pay out.

This distinction matters immensely for the typical reader of this book, who is far more likely to be building a NEPSE portfolio of minority stakes than acquiring companies outright. It means that, for most day-to-day decisions covered in this Canon, the DDM and relative-valuation lenses will often carry more practical weight than a textbook enterprise DCF built as if you were about to buy the whole company and redirect its cash flows at will — though a DCF remains valuable even for minority investing, because it forces explicit thinking about growth, risk, and long-run earnings power that a simple multiple can obscure. The purpose does not eliminate any method from consideration; it changes which method deserves the most weight, and which assumptions (a control premium, a liquidity discount for a minority, illiquid holding) need to be layered on top of the basic mechanics.