Where DCF and growth models break down

What can go wrong?

Two kinds of problem. Bottom-up (company-specific) limitations hit the cash flow: the cash flow measure itself, wrong assumptions or models, and company changes the model cannot capture. Top-down (macroeconomic and market) limitations hit the inputs to the cost of capital in the denominator (the weighted average cost of capital, WACC, and the required return on equity, $r_e$), which move with the economy.

The earlier lessons already built in some cautions:

• Growth $g$ must be below the required return and should not beat the economy forever (lesson 3).
• Constant growth from one year’s cash flow is a simplification.
• The shorter the horizon, the more the terminal value weighs (lesson 5).
• The gap between the required return and $g$ drives the value.

Limitations of DCF growth models

• Bottom-up (company-specific) Memorize the numerator
  • Incomplete measure dividends miss buybacks; payout and share-count shifts
  • Base year one-off years bias growth and forecasts
  • Static assumptions transformation, decline, sudden distress
• Top-down (macro and market) Memorize the denominator
  • Changing risk-free rate moves r_d, r_e and WACC
  • Developing markets no truly risk-free or long-term rate

When the cash flow measure tells only part of the story

Any single measure can bias the value if it misses how the company returns or uses cash. This is the incomplete cash flow metric problem.

Dividends ignore buybacks. A company that keeps a steady payout but returns more and more through buybacks will look too cheap in a dividend model compared with a model based on free cash flow to equity (FCFE). This is the dividend versus FCFE valuation gap .

The gap can also run the other way. Dividends per share can rise quickly while earnings are flat, because the payout ratio is rising or buybacks are shrinking the share count. Growing total dividends at that per-share rate can make a dividend value higher than an FCFE value; the gap comes from the payout ratio, the share count, or both.

An FCFE forecast at a constant debt ratio inherits any oddity in the base year. Operating profit, margins and investment in long-term assets and working capital can be hard to define for both FCFE and free cash flow to the firm (FCFF). These base-year effects also bias historical growth rates of volatile measures (lesson 3).

When the company does not stand still

Growth-rate models suit mature firms whose base-year cash flow is a fair starting point. Two stages allow change within set limits, but not reinvention. These are static growth assumptions .

Kestrel Mowers made petrol lawnmowers for decades. Then it poured its cash and heavy new borrowing into home batteries and a rooftop solar installation business, cut its dividend and sold its old factories. No base-year cash flow and growth rate could describe what it was turning into. That is transformational growth .

Present value models also rest on the going concern assumption . A negative growth rate can model a slow decline (lesson 5), but financial distress (bankruptcy, restructuring, liquidation) often arrives suddenly. It is better modeled as an explicit scenario with a probability. The result is a probability-weighted scenario value .

$$IV_t = p \times IV_{\text{going concern}} + (1 - p) \times \text{Value in distress}$$

Here $p$ is the probability the firm survives.

Learning Module 8 builds full probability-weighted scenarios.

When interest rates move

The risk-free rate sits inside both the cost of debt and the cost of equity, so when it changes, WACC, $r_e$ and the value all move, often as much as any company-specific factor. This is the changing risk-free rate problem.

What analysts do: the practice of matching the risk-free rate to the forecast horizon uses a long-run average rather than today’s short-term rate. Developing markets add a problem: government debt may not be risk-free, and long-term government bond yields may not exist (developing market discount rates ).

So what do analysts do?

Use present value growth models where they fit best: stable firms with consistent dividends and cash flows. Be wary with non-payers, firms with negative cash flow and firms at an early or late turning point in their life cycle. Remember that technology and competition can break static growth assumptions. And practice using more than one model and approach.