Alternative valuation methods

When a standard DCF is the wrong tool: dividend discount for banks and insurers, residual income and NAV for REITs, and sum-of-the-parts for conglomerates, each a repair for a specific way the standard machine breaks.

The standard DCF assumes you can separate operations from financing and define free cash flow cleanly. Whole industries break those assumptions: a bank's debt IS its operations, a REIT's depreciation is fiction, an insurer's costs arrive years after its revenue. Each break has a standard repair, and knowing which tool fits which business is what separates an analyst from a template.

The dividend discount model: the original DCF

The oldest valuation logic there is: a share is worth the dividends it will pay. The workhorse form is the growing perpetuity applied to dividends, usually with a two-stage version for companies still maturing:

value per share = D1 / (r - g)

D1 = next year's dividend    r = cost of equity    g = perpetual dividend growth

two-stage: discount each dividend through the high-growth years explicitly,
then apply the formula to the steady state that follows

The DDM is naive for companies that barely pay out, but it is the natural language for mature, regulated, payout-defined businesses: utilities, telecoms, and, most importantly, banks, where dividends are the one cash flow that verifiably leaves the firm.

Banks: where WACC stops making sense

A bank borrows money as its raw material and lends it as its product. Debt is not financing; it is inventory. That destroys the FCFF/WACC machine twice over: there is no meaningful enterprise value separate from the balance sheet, and "free cash flow" is undefinable when loans are the business. The profession values banks on equity directly:

• Dividend discount, constrained by regulatory capital: the dividends a bank can pay are what it earns minus what regulators make it retain as capital against its risk-weighted assets.
• Residual income / excess returns (below), which prices the equity off the spread between ROE and the cost of equity.
• P/TBV against ROE as the market cross-check: banks earning above their cost of equity trade above tangible book, banks earning below it trade below, and the relationship is close to linear across the sector.

justified P/TBV = (ROE - g) / (r - g)

a bank with 14% ROE, 10% cost of equity, 3% growth:
(0.14 - 0.03) / (0.10 - 0.03) = 1.57x tangible book

Residual income: value as book plus excess earnings

Reframes value instead of forecasting distributions: start from book value, then add the present value of earnings ABOVE the required return on that book. Earning exactly the cost of equity adds nothing; only the excess creates value.

equity value = book value + PV of all future (net income - r x book value)

the term in parentheses is "residual income": profit beyond what
shareholders required for the capital employed

Its virtue is that most of the value sits in the observable book value rather than a distant terminal guess, so the model is far less terminal-dominated than a DCF. It is the academic backbone of bank valuation and the cleanest way to say what ROE-above-cost actually buys.

REITs: earnings lie, so the sector built its own metrics

Real estate depreciates on the books while often appreciating in fact, so a REIT's net income is systematically understated and nobody prices REITs on P/E. The sector's repairs:

FFO  = net income + real estate depreciation - gains on property sales
AFFO = FFO - recurring maintenance capex and leasing costs

FFO is the sector's "earnings"; AFFO is the sector's "free cash flow"
and the honest base for the dividend

• P/FFO and P/AFFO replace P/E as the trading multiples.
• NAV (net asset value): value the property portfolio directly by applying market cap rates to each property's net operating income, subtract debt, and compare the result per share to the stock price. REITs habitually trade at premiums or discounts to NAV, and that gap is the sector's core debate.
• Dividend models work well too, since REITs must pay out 90% of taxable income by law.

property value = net operating income / cap rate

a building earning $10M NOI at a 5% cap rate is priced at $200M;
cap rates move with interest rates, which is why REITs are rate-sensitive

Insurers: costs that arrive years after revenue

An insurer sells a promise and learns its cost of goods sold years later, when claims settle. Earnings therefore depend on reserve estimates, and the sector reads specialized gauges: the combined ratio (claims plus expenses over premiums, under 100% means underwriting at a profit), investment income on the float (premiums held between collection and claims, the engine Buffett built Berkshire on), and valuation on P/B against ROE, same logic as banks. Life insurers add embedded-value methods that present-value the existing policy book.

Resource companies: the asset has a countdown

An oil field or mine depletes; perpetuity math is wrong by construction. The sector standard is NAV by asset: model each field's production profile to exhaustion at forward commodity prices, discount (the convention is a flat 10%, written "PV-10"), sum the assets, subtract corporate costs and debt. Exploration upside rides as a separate, heavily discounted line.

Sum-of-the-parts: when one company is several

Conglomerates and multi-segment companies get valued in pieces: each segment by the method its industry deserves (the streaming arm on revenue multiples, the parks on EBITDA, the stake in a listed subsidiary at market), then summed, netted for corporate costs and debt, and compared to the price. The recurring finding is the conglomerate discount: the market pricing the whole below the sum, which is either the analyst's error or the activist's opportunity, and telling those apart is the work.

Choosing the tool