Discounted Cash Flow (DCF) Valuation

First published 23 Aug 2026 · Last verified 29 Aug 2026

Lesson 46.1 — The Logic of Discounting: Why a Rupee Tomorrow Is Worth Less Than a Rupee Today

Every DCF valuation rests on one simple, almost obvious human intuition: a promise of money in the future is worth less than the same amount of money in hand right now. If a friend offers you a choice between NPR 100,000 today or NPR 100,000 in exactly one year, and you trust the friend completely, you would almost certainly take the money today. Why? Because you could deposit that NPR 100,000 in a fixed deposit at a commercial bank, earn interest on it for a year, and end up with more than NPR 100,000 by the time your friend's promised payment would have arrived. Money today can be put to work; money promised for later cannot be put to work until it arrives. This is the entire idea behind what economists and analysts call the time value of money — the principle that a given sum of money is worth more the sooner it is received, because of its earning potential in the interim.

Discounted Cash Flow (DCF) valuation is nothing more than this everyday intuition applied rigorously, year by year, to an entire business. Instead of asking "what is NPR 100,000 next year worth to me today," a DCF model asks "what is this company's entire stream of future cash — the cash it will generate this year, next year, the year after, and so on into the distant future — worth to an investor today?" To answer that question, an analyst does two things. First, she projects, or forecasts, how much spare cash the business is likely to generate in each of the coming years — this projected number is called free cash flow (FCF), the cash a business has left over after it has paid for its operations and reinvested what it needs to keep running and growing. Second, she converts each future year's projected cash flow into today's rupees using a discount rate — the annual rate of return an investor could reasonably expect to earn elsewhere for taking a similar level of risk, which functions as the "exchange rate" between a rupee tomorrow and a rupee today. The result of applying that discount rate is called the present value of that future cash flow — what a rupee to be received in the future is worth in today's terms once you have accounted for both the passage of time and the risk that the cash might not arrive at all, or might arrive in a different amount than forecast.

Return to the fixed deposit analogy, because it is the cleanest way to understand the mechanics. Suppose a Nepali fixed deposit currently pays 8% per year. If you were promised NPR 108 in exactly one year, and you had access to that 8% deposit rate today, the promise of NPR 108 next year is worth exactly NPR 100 to you today — because NPR 100 placed in the deposit today would itself grow into NPR 108 in one year. The NPR 100 is the present value of the NPR 108 future amount, using 8% as the discount rate. If instead the promise were riskier — say, a promise from a private company rather than a government-backed bank, where there is some chance the money never arrives — you would not be willing to pay NPR 100 today for a promise of NPR 108 in a year. You would want a higher expected return to compensate for the extra risk, meaning you would only pay something less than NPR 100 for that same promised NPR 108. This is why the discount rate used in a DCF is not simply "whatever the bank pays" — it must reflect the specific riskiness of the specific cash flow being valued. A government treasury bill and a small unlisted hydropower company's equity are not the same kind of promise, and they cannot be discounted at the same rate.

The full arithmetic of a DCF, then, is: project free cash flow for each year of an explicit forecast period (commonly five to ten years), discount each year's projected cash flow back to today using the appropriate discount rate, add up all those present values, then add the present value of everything the business is expected to generate beyond the forecast period — a single lump-sum figure called terminal value, which will be explained fully in Lesson 46.4. The sum of all of this — the present value of near-term cash flows plus the present value of the terminal value — is the DCF's estimate of what the business, or a share of it, is intrinsically worth today. This estimate is then compared against the price the market is currently charging for the shares. If the DCF value per share is meaningfully higher than NEPSE's quoted price, the stock may be undervalued; if the DCF value is meaningfully lower, the stock may be overvalued. Chapter 45 already cautioned that "may be" is doing a great deal of work in that sentence — a DCF is an estimate built on assumptions, not a certificate of truth — and this chapter will return to that caution repeatedly, because in Nepal's market the assumptions matter even more than usual.

It is worth pausing on why DCF matters especially for Nepali investors, given how this book has framed NEPSE in earlier chapters. Nepal's stock market is dominated by relative valuation — investors comparing one bank's price-to-book ratio to another's, or one hydropower stock's price-to-earnings ratio to the sector average. Relative valuation is fast and easy, but it only tells you whether something is cheap or expensive relative to its neighbours; it says nothing about whether the whole neighbourhood is fairly priced. DCF is the tool that lets an investor step outside the crowd's current mood and ask a more fundamental question: based on the actual cash this business is likely to generate over its life, and based on what that cash is worth in today's rupees, what should I be willing to pay? It is a slower, more demanding exercise, but it is also the discipline that separates institutional-grade investing from simply following the herd on Maharajgunj coffee-shop tips about which hydropower counter is about to "run."

Lesson 46.2 — Free Cash Flow: The Real Number a DCF Runs On

Before any discounting can happen, an analyst needs a number to discount — and that number is not accounting profit. Net profit, the "bottom line" figure companies report in their income statements and that most Nepali retail investors fixate on, is not the same thing as cash. Net profit can be inflated by non-cash accounting entries, and it does not account for the cash a company must spend on new equipment, new working capital, or debt repayment just to keep functioning and growing. Free cash flow strips all of that away and asks a blunter question: after a company has run its operations and reinvested whatever it needs to reinvest, how much actual, spendable cash is left over for the people who have a claim on it?

There are two versions of free cash flow used in DCF models, and confusing them is one of the most common mistakes analysts make, so it is worth defining both carefully.

Free Cash Flow to the Firm (FCFF) is the cash generated by the business that is available to all providers of capital — both the shareholders (equity holders) and the lenders (debt holders) — before any interest payments or debt repayments are made. It represents the cash the operating business itself throws off, independent of how that business happens to be financed. FCFF is typically calculated starting from operating profit, adding back non-cash charges like depreciation (the accounting expense that spreads the cost of a long-lived asset, such as a hydropower plant's turbines, over its useful life, even though the cash for it was spent upfront), then subtracting capital expenditure (capex — the cash spent on buying, building, or upgrading long-term assets like power plants, factory equipment, or bank branches) and the change in working capital (the cash tied up in day-to-day operating needs like inventory, receivables from customers, and payments owed to suppliers).

Free Cash Flow to Equity (FCFE) is the cash left over for shareholders alone, after the company has already paid its lenders — meaning after interest expense and after any net debt repayments (or plus any net new borrowing). FCFE is what shareholders could theoretically receive as dividends without harming the company's ability to keep operating and growing. FCFE = FCFF − interest expense × (1 − tax rate) − net debt repayments, or, built up directly from net profit, FCFE = Net Profit + Depreciation − Capex − Change in Working Capital + Net Borrowing.

Which approach should a Nepali analyst prefer? For most NEPSE-listed non-financial companies — manufacturers, hotels, hydropower developers, trading companies — either approach can work, but FCFE tends to be more practical for a retail or institutional equity investor because it produces equity value (and therefore value per share) directly, without the extra step of subtracting net debt at the end. FCFE is also more intuitive when a company's capital structure — the mix of debt and equity financing it uses — is expected to stay relatively stable, which is a reasonable assumption for a mature manufacturer but a poor one for, say, a hydropower project still in its debt-heavy construction and early-operation years, where debt is being steadily repaid out of operating cash flow and the capital structure is shifting every year. In that hydropower case, FCFF and WACC can actually be the cleaner approach, because it separates the operating cash-generating power of the plant from the specific, and shifting, debt load sitting on top of it — and only at the very end does the analyst subtract the (declining) net debt to get to equity value.

There is one crucial exception where FCFF/WACC is almost mandatory in Nepal: banks and financial institutions. For a commercial bank or a development bank, debt (in the form of customer deposits) is not really "debt" in the ordinary sense — it is the raw material of the business, not a financing choice layered on top of operations. Interest paid to depositors is an operating cost of banking, not a financing cost to be added back. Trying to build an FCFF for a bank produces a distorted, nearly meaningless number. For banks, DCF practitioners almost universally use a variant built on FCFE directly — often approximated as net profit adjusted for the capital a regulator requires the bank to retain to support its loan book (since Nepal Rastra Bank's capital adequacy requirements effectively force banks to reinvest a portion of profit as regulatory capital rather than distribute it all as dividends). This matters directly for Nepali investors, since banking and financial shares make up one of the largest blocks of NEPSE's market capitalisation.

A second practical wrinkle specific to Nepal deserves mention here: the quality of the historical financial data an analyst starts from. Projecting future free cash flow begins with understanding past free cash flow, built from several years of audited financial statements. Many Nepali companies outside the heavily regulated banking and insurance sectors have historically had inconsistent disclosure quality — working capital movements buried inconsistently across notes, related-party transactions that are not always fully separated out, and depreciation policies that occasionally change between years in ways that are not clearly flagged. An analyst building a DCF for a Nepali manufacturing or trading company should expect to spend real time simply reconstructing a clean, consistent five-year history of FCFF or FCFE before attempting to project even a single year forward. Skipping this reconstruction step and projecting directly off a single reported "net profit" figure is one of the fastest ways to build a DCF that looks precise but is quietly wrong.

Lesson 46.3 — Building the Discount Rate: Cost of Equity in a Frontier Market

If free cash flow is the numerator of a DCF, the discount rate is the denominator, and in a market like Nepal's, the denominator is where most of the genuine difficulty — and most of the genuine judgment — lives. The standard tool for estimating the cost of equity (the annual return shareholders require to compensate them for the risk of owning a particular stock, as opposed to the cost of debt, which is simply the interest rate lenders charge) is the Capital Asset Pricing Model (CAPM). CAPM states:

Cost of Equity = Risk-Free Rate + Beta × Equity Risk Premium

Each of these three pieces requires real judgment when applied to a NEPSE-listed company, so each is worth walking through carefully.

The risk-free rate is the return available on an investment with effectively no default risk — conventionally the yield on a government security, since a government that can print its own currency is assumed (rightly or wrongly) never to default on debt issued in that currency. In Nepal, the natural proxies are yields on Government of Nepal Treasury Bills (short-term instruments with maturities under a year, auctioned regularly by Nepal Rastra Bank on behalf of the government) and Development Bonds (longer-maturity government securities). The trouble is that these yields have been remarkably volatile in Nepal over the past several years — not because Nepal's government credit risk has changed dramatically, but because of swings in domestic banking-sector liquidity. When Nepali banks are flush with deposits and have limited lending opportunities, they pile into treasury bills, driving yields down sharply; when liquidity tightens, yields spike. Nepal's weighted average treasury bill rate swung from double digits during the 2022 liquidity crunch down to a much lower single-digit range during the liquidity glut of 2024 and 2025, before partially normalising. A "risk-free rate" that can move by five or more percentage points within two years, for reasons having nothing to do with Nepal's actual sovereign creditworthiness, is not a clean input.

The equity risk premium (ERP) is the extra return, above the risk-free rate, that investors as a whole demand for holding a diversified basket of equities rather than the risk-free asset. It compensates for the general riskiness of stocks as an asset class. Mature-market ERPs (built from many decades of, for example, US or developed-market stock market history) are commonly estimated in a broad range around 4–6%. Nepal is not a mature market, and NEPSE has nowhere near the multi-decade, high-quality return history that underpins those mature-market estimates. Practitioners valuing companies in frontier and emerging markets typically address this by starting with a mature-market ERP and adding a country risk premium — an additional increment reflecting the extra risk of investing in a specific country's equity market, often estimated (following the widely used approach popularized by NYU professor Aswath Damodaran) by taking that country's sovereign credit rating or bond default spread and scaling it up by the relative volatility of that country's equity market compared to its bond market. For a market like Nepal's — unrated or low-rated by major international agencies, with capital controls restricting foreign portfolio investment and comparatively low market liquidity — the resulting country risk premium is substantial, often adding several percentage points on top of a mature-market ERP base. The practical upshot is that a Nepali company's cost of equity, built up honestly through this framework, tends to land meaningfully higher than a comparable company's cost of equity in a developed market — frequently in the low-to-mid teens as a percentage, even before considering the specific riskiness of an individual firm.

Beta — the third component of CAPM — measures how much a specific stock's returns move relative to the overall market's returns; a beta of 1.0 means the stock tends to move in line with the market, a beta above 1.0 means it tends to amplify market moves, and a beta below 1.0 means it tends to be more stable than the market. In deep, liquid markets, beta is estimated by running a regression of a stock's historical returns against the broader index's returns over several years of daily or weekly data. NEPSE presents a specific and serious problem here: a large share of listed stocks trade thinly, some going days or weeks without a single transaction, and the exchange's daily circuit breakers (price movement limits that halt or cap how far a stock can move in a single session) further compress and distort the reported price series. A beta regression run on a NEPSE stock's historical prices is regressing against data that often does not reflect genuine, continuously-clearing market prices — it reflects whatever price happened to clear on the handful of days a trade actually occurred, subject to circuit-breaker ceilings and floors. The result is that raw, statistically-estimated betas for many NEPSE stocks are unreliable: they can appear artificially low (because a stock that barely trades appears to not move much, understating its true risk) or erratic from one estimation period to the next.

Given this, practical Nepali analysts commonly lean on one or both of two workarounds. The first is a bottom-up beta approach: rather than regressing a specific Nepali company's own erratic price history, the analyst starts from the beta of comparable listed companies in the same industry from markets with better data (for example, listed hydropower or manufacturing companies in India or other South Asian markets with deeper trading histories), "unlevers" those betas to strip out the effect of each comparable company's own debt load (since more debt mechanically amplifies equity risk), averages the unlevered figure across several comparables to get an industry-level unlevered beta, and then "relevers" that industry beta using the specific Nepali company's own capital structure (its own mix of debt and equity) to arrive at a beta tailored to that company. The second workaround is simply to use a small set of reasonable qualitative brackets — for example, treating regulated, essential-service businesses with stable demand (established commercial banks, hydropower plants with long-term power purchase agreements already signed and in operation) as lower-beta (perhaps 0.7–0.9), and treating more cyclical, discretionary, or leverage-heavy businesses (hotels and tourism, construction, early-stage hydropower still exposed to hydrology and construction risk) as higher-beta (perhaps 1.1–1.4), and defending that judgment explicitly in the write-up rather than hiding behind a spuriously precise regression output.

Putting the pieces together with illustrative, order-of-magnitude figures: a risk-free rate proxy of roughly 7% (a normalised, multi-quarter average of longer-maturity Nepali government securities, deliberately smoothing over the recent liquidity-driven swings), a mature-market ERP of roughly 5%, a Nepal country risk premium of roughly 4–5% given the market's frontier status, and a company-specific beta of, say, 1.0 for a mid-risk operating business, would combine to a cost of equity in the neighbourhood of 16–17%. A lower-risk regulated utility with beta 0.8 might land closer to 14–15%; a higher-risk, highly leveraged cyclical business with beta 1.3 might land at 18% or more. These are illustrative ranges, not fixed rules — the discipline is in building the number up transparently from its components, stating every assumption explicitly, and then stress-testing the final valuation against a plausible range of discount rates (a topic taken up fully in Lesson 46.6), rather than presenting a single false-precision figure like "14.73%" as though it were an observed fact rather than a constructed judgment.

For FCFF-based valuations, the discount rate is WACC rather than the cost of equity alone — a blend of the cost of equity and the after-tax cost of debt (the interest rate a company actually pays lenders, adjusted downward for the tax shield since interest expense is tax-deductible), weighted by the company's proportion of equity and debt financing at market value. Estimating the cost of debt for a Nepali company is comparatively more straightforward, since it can usually be observed directly from the interest rates the company actually pays on its bank loans and debentures, as disclosed in its financial statements — though for heavily-indebted hydropower projects, this figure deserves its own scrutiny, since concessional or subsidized lending rates on some hydropower debt (occasionally available through targeted refinancing facilities) can understate what the company would pay to raise fresh debt today.

Lesson 46.4 — Terminal Value: Valuing the Business Beyond the Forecast Horizon

No analyst can credibly forecast a company's cash flows line by line, year by year, forever. Most DCF models therefore build an explicit forecast for a limited number of years — commonly five, sometimes as many as ten for a business still ramping toward a stable, mature state, such as a hydropower plant in its early years of operation — and then collapse everything the business is expected to generate beyond that horizon into a single figure called terminal value, calculated as of the last year of the explicit forecast and then discounted back to today just like any other cash flow.

There are two standard methods for calculating terminal value, and a careful analyst should ideally compute both and compare them as a sanity check on each other.

The Gordon Growth Model (also called the perpetuity growth method) assumes that, after the explicit forecast period ends, the company's free cash flow will grow at a constant, modest rate forever. The formula is:

Terminal Value = FCF in the first year after the forecast period ÷ (Discount Rate − Perpetual Growth Rate)

The perpetual growth rate used here must be conservative and sustainable — a rate no company can realistically exceed forever, since growing faster than the overall economy indefinitely would eventually mean the company becomes larger than the economy itself, which is impossible. In practice, this means the terminal growth rate should be anchored near a country's expected long-run nominal GDP growth rate (real GDP growth plus inflation) — for Nepal, taking into account the country's growth trajectory and inflation history, a terminal growth rate in the region of 4–6% is a commonly defensible anchor, though this should be revisited as Nepal's macroeconomic outlook evolves. Using a terminal growth rate materially above this — say, 8% or 10%, forever — is one of the most common ways analysts (deliberately or accidentally) inflate a DCF valuation, because the terminal value is extremely sensitive to the gap between the discount rate and the growth rate, as the worked example in the next lesson will demonstrate numerically.

The Exit Multiple Method instead assumes that, at the end of the explicit forecast period, the business could be sold for a price based on some valuation multiple (such as EV/EBITDA — enterprise value divided by earnings before interest, tax, depreciation, and amortisation, a common proxy for operating cash-generating power — or a price-to-earnings multiple) observed from comparable companies at that future point in time. This method has the advantage of grounding the terminal value in something closer to observable market pricing, but it has a specific weakness in the Nepali context: it requires a reasonable set of comparable companies with credible, liquid market pricing, and NEPSE's universe of comparables within any single narrow sub-sector (a handful of hydropower developers, a handful of hotels, a cluster of similarly-sized commercial banks) is small, and their trading multiples are themselves influenced by the same illiquidity and sentiment-driven swings discussed throughout this book. An exit multiple pulled from a thinly-traded, sentiment-driven peer group can smuggle exactly the kind of market mispricing a DCF is supposed to help an investor see past back into the "intrinsic" valuation, defeating much of the purpose of doing a DCF in the first place.

This concentration of value in the terminal period has a specific implication for how DCF should be used on Nepali cyclical and hydropower-type businesses discussed in earlier chapters of this book. A hydropower plant's cash flow is not a smooth, steadily growing line — it depends heavily on hydrology (the seasonal pattern of river flow, with Nepal's rivers running high during the monsoon months of roughly Shrawan to Ashwin and much lower during the dry winter and pre-monsoon months), on the specific tariff structure locked in under its Power Purchase Agreement (PPA) with the Nepal Electricity Authority (the near-monopoly state utility that buys power from the large majority of Nepal's independent hydropower producers, typically under long-term, fixed or seasonally-differentiated tariff contracts), and on when its underlying debt is scheduled to be repaid. A well-built hydropower DCF should model these mechanics explicitly year by year through the explicit forecast period — rising cash flow as construction-linked debt is paid down, seasonal swings between wet-season and dry-season generation reflected in the tariff structure, and a step-change around the point (often twenty to thirty years from commissioning, depending on the specific PPA and license terms) when the PPA tariff structure or the operating license itself may reset or expire — rather than smoothing all of that complexity into one constant terminal growth rate applied indefinitely. For a hydropower asset with a licensed life that is finite (Nepali hydropower licenses are typically issued for a fixed number of years, after which ownership arrangements can change), a straightforward Gordon Growth terminal value assuming cash flows forever is not merely imprecise — it can be conceptually wrong, and a more careful model may need to explicitly value only the remaining licensed life plus a defensible view (rather than an assumption) on what happens at license expiry.

Lesson 46.5 — A Worked Example: Discounting a Nepali Company Step by Step

The mechanics described so far are easiest to absorb through a complete numerical example. Consider a hypothetical company, Himalaya Hydro Power Ltd. (HHPL) — a fictional, mid-sized run-of-river hydropower company, illustrative of the kind of counter that trades on NEPSE's hydropower sub-index, with 3 crore shares outstanding and a long-term PPA already signed with the Nepal Electricity Authority. HHPL's plant has been operating for several years, its construction-period debt is being steadily repaid, and its free cash flow to equity (FCFE) is expected to grow as debt service costs fall each year, before settling into a stable, modest long-run growth pattern once the debt is largely repaid and generation output stabilises.

Suppose an analyst has done the work described in Lessons 46.2 and 46.3: reconstructed HHPL's historical FCFE from several years of audited statements, projected FCFE forward five years reflecting the declining debt-service burden, and built up a cost of equity of 14.5% using the CAPM framework — a risk-free rate proxy around 7%, a combined equity and country risk premium reflecting Nepal's frontier-market status, and a beta near 1.0 reflecting HHPL's moderate operating and financial leverage. The analyst has also settled on a terminal growth rate of 4%, anchored to a conservative view of Nepal's long-run nominal GDP growth.

Reading this table mechanically: each year's projected FCFE is multiplied by its discount factor (1 ÷ (1.145)^n for year n) to arrive at that year's present value — for instance, Year 3's NPR 18 crore, five years before it lands in HHPL's shareholders' hands in full, is worth NPR 12.00 crore in today's terms once discounted at 14.5% for three years. The five years of explicit present values sum to NPR 56.24 crore. Terminal value is then calculated using the Gordon Growth formula on the cash flow expected in the first year after the forecast window: Year 5's FCFE of NPR 21 crore, grown one more year at the 4% terminal rate, gives NPR 21.84 crore, divided by the gap between the discount rate and the growth rate (14.5% − 4% = 10.5%), giving a terminal value of NPR 208.00 crore as of the end of Year 5. That terminal value is itself a future amount as of today, so it must also be discounted back five years at 14.5%, giving a present value of NPR 105.77 crore. Adding the present value of the explicit forecast period (NPR 56.24 crore) to the present value of the terminal value (NPR 105.77 crore) gives a total estimated equity value of NPR 162.01 crore. Dividing by HHPL's 3 crore outstanding shares gives an estimated intrinsic value of roughly NPR 54.00 per share.

Notice immediately, as flagged in Lesson 46.4, that the terminal value's present value (NPR 105.77 crore) makes up about 65% of HHPL's total estimated equity value — the majority of what this valuation claims the company is worth rests on a single assumption about growth forever after Year 5, discounted back through a single assumption about the appropriate long-run discount rate. This is not a flaw specific to this example; it is a structural feature of DCF valuation applied to almost any going concern, and it is precisely why Lesson 46.6 turns next to testing how sensitive this NPR 54.00-per-share estimate actually is to reasonable changes in those two assumptions.

Lesson 46.6 — Sensitivity, Fragility, and the Practical Limits of DCF in Nepal

The HHPL example above produced a single, clean number: NPR 54.00 per share. Presented alone, that number carries an unearned air of precision — as though the analyst has discovered a hidden truth about what HHPL is really worth. A properly disciplined DCF exercise never stops at a single point estimate; it immediately asks how much that estimate moves when the two most influential assumptions — the discount rate and the terminal growth rate — are varied within a reasonable range. This exercise is called sensitivity analysis, and it is arguably more informative than the base-case number itself, because it reveals how much of the conclusion is really being driven by the analyst's judgment calls rather than by the underlying business.

Holding HHPL's projected FCFE stream fixed and varying only the discount rate and terminal growth rate produces the following range of per-share values:

The spread here is striking: across a discount rate range of only four percentage points (12.5% to 16.5%, both defensible depending on how one estimates beta and the country risk premium) and a terminal growth range of only four percentage points (2% to 6%, both defensible depending on one's view of Nepal's long-run nominal growth), HHPL's estimated per-share value swings from roughly NPR 40.7 to roughly NPR 83.1 — more than double from the low end to the high end. An investor who took only the base-case NPR 54.00 figure at face value, without appreciating this range, could be lulled into false confidence about the precision of the estimate. The honest and useful way to present a DCF output is therefore not "HHPL is worth NPR 54.00 per share" but rather "under a set of reasonable assumptions, HHPL appears to be worth somewhere in the range of roughly NPR 40 to NPR 85 per share, with NPR 54 as a central estimate" — and then to compare that full range, not just the midpoint, against NEPSE's quoted market price to judge whether the stock looks cheap, expensive, or fairly valued.

Beyond the mechanical sensitivity of discount rate and terminal growth, several practical challenges specific to Nepal deserve a final, direct discussion, because they affect how much weight a Nepali investor should place on any DCF output at all.

The first is data quality, already touched on in Lesson 46.2. A DCF is a forward-looking exercise built on a backward-looking foundation — the analyst must first understand several years of a company's true, cash-based operating history before projecting it forward, and for many NEPSE-listed companies outside banking and insurance, that historical foundation is genuinely harder to build cleanly than it would be for a comparable company in a market with longer-established, more consistently enforced disclosure and audit standards.

The second is currency and inflation. Nepal has experienced periods of meaningfully higher inflation than the developed-market economies whose DCF textbooks and default assumptions much of standard valuation theory is built around, and the Nepali rupee's value, while pegged to the Indian rupee, is exposed indirectly to broader currency and external-sector pressures (import dependence, remittance inflows, and foreign exchange reserve levels, all discussed elsewhere in this book). A DCF must be built consistently in either nominal terms (cash flows and discount rate both including expected inflation) or real terms (both excluding it) — mixing the two, for instance by discounting inflation-adjusted "real" cash flow projections at a nominal, inflation-inclusive discount rate, is a subtle but common error that silently and substantially understates value, and it is worth double-checking explicitly in any Nepali DCF given how much inflation assumptions have moved over the past several years.

The third, most acute for the specific businesses this book has covered in earlier chapters, is the genuine difficulty of forecasting cash flow for cyclical and project-based businesses. A hydropower developer's cash flow depends on hydrology that can vary meaningfully from year to year, on monsoon timing, and on the specific terms of a PPA that may not be fully public or fully understood by outside analysts. A hotel or tourism business's cash flow depends on tourist arrival cycles that are themselves sensitive to regional geopolitics, global travel patterns, and domestic political stability — none of which lend themselves to confident five-year-ahead point forecasts. A trading or manufacturing company's cash flow can be exposed to volatile import costs, exchange rate pass-through, and shifting government duty structures. In each of these cases, a DCF is still a useful discipline — it forces an analyst to think explicitly, year by year, about what actually drives the cash — but the honest output is a range of plausible values under different scenarios (a strong monsoon year versus a weak one, a tourism recovery scenario versus a stagnation scenario), not a single confident number.

None of this means DCF should be abandoned in favour of purely relative, multiple-based valuation on NEPSE. It means DCF should be used the way a careful engineer uses a calculation with known error bars, rather than the way a fortune-teller uses a crystal ball: as a disciplined way of making assumptions explicit, testing how much each assumption matters, and forming a reasoned view of value — always held with appropriate humility about how much of the final number is genuine insight into the business and how much is simply the analyst's own assumptions echoing back.