Module I· PP&E, Capex & D&AAdvanced
Question
How do you model the capex cycle for a cyclical industrial — the typical mistakes?
Answer
Mechanics
Cyclical industrials (steel, chemicals, auto suppliers) show huge capex swings over the economic cycle — in boom years capacity expansions are caught up on, in downturn years capex is cut to a minimum.
Typical DCF mistakes
- Extrapolating capex from 1 'normal' base year — either a boom year (overstated) or a recession year (understated); valuation distorted by 15–30%.
- Forecasting D&A on a straight line with no capex lag — in capex boom years D&A is too low, in later years too high.
Clean approaches
- Mid-cycle capex: a 7–10-year average as the steady state.
- Normalize capex / sales to a 7–10-yr average — typically 4–8% at cyclical middle-market companies.
- Capex phasing in the DCF: years 1–2 reflect the current cycle, years 3+ converge to mid-cycle.
- Terminal capex = terminal D&A for steady-state conditions.
Deep diveShow more details
Example
A large cyclical steelmaker — capex swings between 200 (downturns) and 600 (booms), mid-cycle ~400. A DCF built on a single year produces systematically wrong valuations.
Pitch tip
'The DCF is based on mid-cycle capex of $400m; the sensitivity shows ±15% equity value across a capex range of $350–450m'.