Discounted cash flow: why future money is worth less today

This matters because most equity valuations are not built on the current year. They are built on years that have not happened. Sometimes they are built mainly on a terminal value that sits beyond the explicit forecast period and can represent 60% to 80% of total enterprise value. That is where many DCF models stop being analysis and become disguised optimism.

The mechanics of time value: why today’s dollar gets seniority

The dcf time value of money principle is simple. A dollar today can be invested. A dollar next year cannot be used until next year. The difference is not philosophical. It is arithmetic.

If an investor can earn 8% on capital elsewhere, then $100 received one year from now is not worth $100 today. It is worth roughly $92.59 today:

$100 / (1 + 0.08) = $92.59

Two years out, the haircut compounds:

$100 / (1.08)^2 = $85.73

By year ten, that same $100 is worth only $46.32 at a 8% discount rate. The business may still receive the cash. The shareholder buying the stock today should not pay a full dollar for it.

That is the central point often lost in casual “long-term compounder” language. Future cash flows do not arrive in a vacuum. They arrive after competition, reinvestment, dilution, capital costs, taxes, cyclicality, and management error have had their turn.

A DCF does not make future cash certain. It merely prices the uncertainty with a discount rate.

This is why discounted cash flow explained properly starts with skepticism, not growth. The first question is not “how large can revenue become?” The first question is “how much cash can be taken out of the business, and how long must the investor wait for it?”

Accounting earnings are only the beginning. Net income includes accruals. It can be helped by capitalized costs, stretched payables, stock-based compensation treatment, and depreciation schedules that do not match economic wear. Free cash flow is harder to flatter for long, though not impossible. Working capital timing can still distort a single year. Capital expenditure can be deferred. Receivables can swell quietly.

A competent DCF therefore starts below the headline earnings line. It asks whether profit converts into cash. If the cash conversion cycle is deteriorating while revenue grows, the model should not reward the growth without penalty. Revenue that must be financed through larger inventories and slower collections is not the same as revenue that falls cleanly into operating cash flow.

Deconstructing the DCF formula and the role of WACC

The basic DCF formula is direct:

DCF = Sum of [CF_t / (1 + r)^t]

Where:

• CF_t is the cash flow in a given period.
• r is the discount rate.
• t is the period number.

The formula is clean. The inputs are not.

A typical company DCF estimates free cash flow for a defined forecast period, usually five to ten years, then discounts those cash flows back to today. The analyst then adds a terminal value, also discounted back to today. Debt, cash, and other claims are adjusted to move from enterprise value to equity value.

The discount rate is commonly the weighted average cost of capital, or WACC. WACC blends the cost of equity and the after-tax cost of debt according to the company’s capital structure. Cost of equity is often estimated through the Capital Asset Pricing Model. Debt is cheaper after taxes because interest is tax-deductible, but debt is not free. It adds fixed claims and bankruptcy risk.

The common structure looks like this:

The role of WACC is not decorative. It decides how aggressively future cash flows are penalized. A higher discount rate reduces the present value of distant cash flows exponentially. This is not linear. It is the reason growth equities can move violently when rates rise or risk premiums widen.

A public SaaS company might be modeled with a baseline discount rate around 10% if it has scale, retention evidence, and a defensible path to cash margins. An earlier-stage startup may require 15% to 20%, or more, because the cash flows are less proven and farther away. For private companies, valuation experts often apply a liquidity premium of 20% to 30% to the WACC, or use a cost of equity in the 25% to 40% range. That is not punishment. It is recognition that private equity cannot be exited at a quoted market price at 9:31 a.m.

The analyst should not pretend there is one sacred discount rate. WACC is common for company-level enterprise valuation. Cost of equity may be more appropriate when valuing equity cash flows directly. A project-specific hurdle rate may be more defensible for a risky internal investment. The rate must fit the claim being valued.

This is also where Net Present Value differs from DCF. A DCF calculation gives the present value of expected future cash flows. NPV subtracts the initial investment cost from that present value. If the present value is $120 million and the required investment is $100 million, the NPV is $20 million. The distinction is small in wording and large in capital allocation.

The terminal value trap: where most of the valuation hides

The explicit forecast period gets most of the spreadsheet attention. Revenue growth by year. Margin progression. Sales efficiency. Tax rates. Capital expenditure. Working capital. Ten tabs of apparent rigor.

Then one cell determines most of the value.

Terminal value estimates what the business is worth beyond the explicit forecast period. In many DCF models, terminal value accounts for 60% to 80% of total enterprise value. That range is not an accounting quirk. It is the ordinary consequence of valuing a continuing business. But it creates a problem: the least knowable part of the model often carries the largest weight.

There are two primary methods.

The first is the Gordon Growth Model, or perpetual growth method. It assumes free cash flow grows at a stable rate forever:

Terminal value = FCF_(n+1) / (r - g)

Here, g is the perpetual growth rate. For a mature U.S. business, a defensible perpetual growth rate is often around 2.0% to 2.5%, roughly aligned with long-run nominal GDP growth. A company cannot outgrow the economy forever without eventually becoming the economy. That observation is not elegant. It is just necessary.

The second method is the exit multiple. The analyst applies a valuation multiple, often EV/EBITDA or EV/free cash flow, to the final forecast year. This looks market-based. It is still a forecast. If the exit multiple is borrowed from inflated peer valuations, the DCF has merely laundered a market multiple through a spreadsheet.

A simple example shows the problem.

Assume a company is expected to produce $100 million of free cash flow in year five. The discount rate is 10%. The perpetual growth rate is 2.5%.

Terminal value at year five:

$100 million × 1.025 / (0.10 - 0.025) = $1.367 billion

Discounted back five years:

$1.367 billion / (1.10)^5 = about $849 million

Now change the perpetual growth rate from 2.5% to 3.0%:

$100 million × 1.03 / (0.10 - 0.03) = $1.471 billion

Discounted back:

$1.471 billion / (1.10)^5 = about $913 million

Half a percentage point adds about $64 million in present value. No factory was built. No customer paid. One assumption moved.

Terminal value is where a DCF most often stops measuring the business and starts measuring the analyst’s tolerance for fantasy.

The same sensitivity appears in the discount rate. At 9% instead of 10%, the valuation expands. At 11%, it compresses. This effect is severe for businesses whose cash flows are expected far in the future. A utility with current cash generation is less sensitive. A software company promising future operating leverage is more sensitive. A biotech with no commercial revenue is mostly sensitivity analysis wearing a lab coat.

Free cash flow is not net income with better branding

The phrase “discounted cash flow means” future cash flows are translated into present value. The word cash deserves attention. A DCF built mechanically from net income can miss the economics.

Free cash flow usually starts with operating cash flow and subtracts capital expenditure. For valuation, the analyst may use unlevered free cash flow, which is cash available to all capital providers before debt payments. That keeps financing decisions separate from operating value. Levered free cash flow, by contrast, is cash available to equity holders after debt service. Both can be valid. Mixing them with the wrong discount rate is not.

The forensic work sits in adjustments.

Stock-based compensation is a common example. Companies often ask investors to ignore it in adjusted earnings. The cash did not leave the company immediately, but shareholders paid through dilution. A DCF that adds back stock-based compensation without modeling dilution or future repurchase needs is overstating owner cash flow.

Capitalized software is another example. Some companies capitalize product development costs, moving expenses from the income statement to the balance sheet. That can lift current operating profit. It does not eliminate economic cost. If capitalized costs are recurring, they should be treated as reinvestment, not as a one-time accounting footnote.

Working capital can be equally blunt. If accounts receivable grow faster than sales, reported revenue may be running ahead of cash collection. If inventory builds, margin may be hiding future markdowns or demand errors. If payables stretch, cash flow may look better because suppliers are financing the company. That is not durable free cash flow. It is timing.

A practical DCF should therefore examine:

1. Cash conversion over several years. One year can be distorted by tax timing, inventory corrections, or collection cycles. A five-year pattern is harder to fake.

2. Maintenance versus growth capital expenditure. Total capex may include expansion. But maintenance capex cannot be ignored. A business that underinvests can flatter current free cash flow and impair future competitiveness.

3. Accruals and revenue quality. Rising accruals relative to earnings deserve a lower confidence level. Cash earnings should carry more weight than accounting earnings.

4. Margin assumptions versus reinvestment needs. Expanding margins while also assuming durable high growth can double-count efficiency. Growth usually consumes something: sales expense, R&D, working capital, or capex.

5. Tax normalization. Temporary tax benefits should not be capitalized as if permanent. The model should reflect a steady-state tax burden unless there is a clear structural reason otherwise.

These are not cosmetic refinements. They decide whether the cash flow line is investable or merely reported.

Why use discounted cash flow when multiples are faster?

Multiples are useful. They are also blunt instruments. A price-to-earnings ratio can screen a market quickly. EV/EBITDA can compare capital-intensive companies while neutralizing some financing differences. Free cash flow yield can highlight dislocations.

But a multiple rarely explains itself. A stock at 18 times earnings may be cheap if earnings are depressed and cash conversion is strong. It may be expensive if margins are cyclically inflated and capex has been deferred. The multiple is an output of expectations, not a substitute for them.

DCF forces those expectations into view. Growth must be typed into the model. Margins must be defended. Reinvestment must be funded. Terminal value must be justified. The method is useful because it exposes the argument.

That does not make it superior in every case. For banks, book value and return on equity often carry more direct relevance because the business model is balance-sheet-driven. For early-stage companies with negative cash flow, DCF can become a long corridor of assumptions. For commodity producers, mid-cycle pricing may matter more than the last reported cash flow. For platform businesses, including trading and brokerage infrastructure, changes in user behavior and regulation can matter as much as near-term monetization; broader context on FinTech industry trends for consumers and businesses can help frame those demand and distribution shifts without replacing valuation work.

The better comparison is not DCF versus multiples. It is internal valuation versus external shorthand.

The DCF’s advantage is discipline. Its weakness is false precision. The output might say $42.37 per share. The honest conclusion is usually a range.

Adjusting for reality: public liquidity and private company risk

A public company valuation benefits from observable prices, reported filings, and some market liquidity. None of those guarantees accuracy. They do reduce certain risks.

Private companies require harsher treatment. Their shares are illiquid. Financial reporting may be less transparent. Customer concentration can be higher. Founder dependency may be material. Debt terms may be less favorable. Exit timing may be uncertain.

That is why private company valuations often use a liquidity premium of 20% to 30% added to the WACC, or a cost of equity in the 25% to 40% range. These numbers are not ornaments for valuation reports. They reflect the fact that capital trapped in an illiquid asset needs a higher required return.

The effect is severe.

A $10 million cash flow ten years from now discounted at 10% is worth about $3.86 million today. At 25%, it is worth about $1.07 million. At 40%, it is worth about $346,000.

Same cash flow. Different risk and liquidity. Different present value.

This is the part many private market marks resist. A company can be promising and still be worth far less today than its future revenue story implies. The discount rate is the market’s objection written in numerical form.

The same issue appears in public small caps, though less formally. Thin trading, customer concentration, weak disclosure, and fragile financing should not receive the same discount rate as a mature large-cap compounder. Liquidity is not binary. It sits on a spectrum. The model should reflect that.

Sensitivity analysis: the model should confess its weak points

A DCF with one output is incomplete. The business is not a bond with fixed coupons. Even bonds have credit risk. Equity cash flows are residual claims after everyone else has been paid.

Sensitivity analysis shows how valuation changes when key assumptions move. The usual variables are discount rate, perpetual growth rate, revenue growth, operating margin, and reinvestment intensity.

A sober model does not hide this. It displays it.

For example, assume a base case value of $50 per share. If the discount rate moves from 9% to 11%, and the perpetual growth rate moves from 3% to 2%, the value may fall into the high $30s. If margins fail to expand as forecast, it may fall further. If cash conversion weakens, the equity value may not merely decline; the whole thesis may break.

This is not a defect. It is the point. DCF valuation meaning lies in mapping the range of plausible values under explicit assumptions.

The useful questions are:

• How much of the value depends on cash flows after year five?
• What margin level is required to justify the current market price?
• Does the company have a record of converting revenue into free cash flow?
• Are reinvestment needs understated?
• Does the terminal growth rate exceed what a mature company can plausibly sustain?
• Is the discount rate compensating for balance sheet, cyclicality, customer concentration, and liquidity risk?

Reverse-engineering the current share price is often more revealing than building a heroic base case. If a stock trades at $80 and conservative assumptions produce $45, the market is underwriting either faster growth, higher margins, lower risk, or a richer terminal multiple. The analyst’s job is to identify which assumption is doing the work.

Sometimes the answer is all of them. That is usually not a bargain.

A sober intrinsic value estimate is a range, not a verdict

Discounted cash flow means future money is discounted into present value because time and risk have cost. That definition is basic. The implications are not.

A DCF model is most useful when it is treated as an audit of expectations. It should reconcile earnings to cash. It should penalize distant and uncertain cash flows. It should make terminal value visible, not bury it below the forecast. It should use a discount rate that matches the risk of the claim being valued. It should admit sensitivity instead of pretending precision.

The final number should usually be expressed as a range. For a mature public company with stable cash conversion, perhaps that range is narrow. For an early-stage company or a private business, it should be wide. If terminal value supplies 70% of enterprise value and the company has no durable record of free cash flow, the range should be wider still.

Intrinsic value is not discovered by typing a perpetual growth rate into a cell. It is estimated by stripping away the parts of reported performance that do not convert into owner cash, then discounting what remains at a rate that respects risk. The result may be uncomfortable. That is not a flaw in the method. It is often the only useful thing the model says.