Introduction (3-4): Investment-appraisal projections cover 5-20+ year horizons during which many variables will change — demand, costs, competition, regulation, technology, macro conditions. 'False precision' is a real risk. The question is how firms should handle this uncertainty responsibly.
Sources of uncertainty (6-7):
(1) Demand uncertainty. Will customers want the product? How much? At what price? For how long?
(2) Cost uncertainty. Raw materials, labour, energy, logistics — all volatile.
(3) Competitive uncertainty. Will competitors enter? Will they undercut? Will substitute products emerge?
(4) Technology uncertainty. Will the technology work? Will better technology emerge mid-project?
(5) Regulatory uncertainty. Will governments change rules? Taxes? Subsidies?
(6) Macroeconomic uncertainty. Recession, inflation, interest rates, currency.
(7) Timing uncertainty. When will investment pay back? Delays extend recovery.
(8) Residual-value uncertainty. What will the asset be worth at end-of-project?
Methods for handling uncertainty (10-11):
(1) Sensitivity analysis. Vary each key assumption ±10-30% and see NPV impact. Reveals which assumptions matter most.
(2) Scenario analysis. Build base / optimistic / pessimistic / severe scenarios combining multiple variables; show NPV in each. Captures correlations missed by single-variable sensitivity.
(3) Monte Carlo simulation. Use probability distributions for key inputs; run thousands of simulations; produce probability-weighted NPV. More sophisticated; requires good input distributions.
(4) Decision trees. Model projects with multiple paths (success / failure, further investment decisions); evaluate expected value.
(5) Real-options analysis. Treat flexibility as valuable — the option to expand / abandon / defer has economic value beyond basic NPV. Captures strategic flexibility.
(6) Conservative base assumptions. Build models on realistic (not optimistic) assumptions; use demand / price estimates in bottom-two-thirds of possible range.
(7) Stage-gate approach. Invest in phases with go / no-go decisions; preserve option to abandon at each stage.
(8) Shorter projection horizons + residual-value approach. Model 5 years explicitly + terminal value — reduces compounding uncertainty.
(9) Qualitative risk-register. List risks with probability and impact; include mitigation plans.
(10) Robust-decision analysis. Choose investments robust across many scenarios rather than optimised for one scenario.
(11) Monitor and course-correct. Build in regular review; adjust investment pace based on how reality unfolds.
(12) Diversify. Portfolio of smaller investments rather than one large bet.
Evaluation (3-4): Modern investment appraisal has multiple tools for handling uncertainty:
(1) Single-point NPV estimates are inadequate — misleadingly precise.
(2) Scenario + sensitivity + Monte Carlo provide richer picture.
(3) Real options capture flexibility value traditional appraisal misses.
(4) Stage-gate turns one big bet into series of smaller informed bets.
(5) Conservative baselines reduce false precision.
(6) Qualitative judgement about residual uncertainty always required.
(7) Firms with sophisticated appraisal (consultancies, large corporates, private equity) use these tools; smaller firms often don't.
Final judgement: firms should use multiple uncertainty-handling tools rather than single-point estimates:
(1) Base case + scenarios + sensitivity for all major investments.
(2) Monte Carlo for high-stakes investments.
(3) Real options for projects with embedded flexibility.
(4) Stage-gate to preserve optionality.
(5) Conservative base assumptions.
(6) Robust-decision principle — prefer projects that work across scenarios over those optimised for one.
(7) Build organisational capability — appraisal quality separates good from bad investment decisions.
'In a volatile world, false precision in investment appraisal is more dangerous than acknowledged uncertainty. The sophisticated firm uses multiple tools, keeps assumptions conservative, preserves optionality, and reviews frequently. The firms that invested confidently with single-point estimates in 2019 for Covid-era projects paid dearly; firms with scenario-based appraisal and real options handled the shock better.' [20]