expected-monetary-value

# Expected Monetary Value (EMV) **Canonical Source**: Decision Analysis / PMBOK Guide (PMI) **Domain**: Project Management, Risk Analysis, Decision Theory **Standard**: PMI PMBOK Guide - Quantitative Risk Analysis ## One-Line Summary Probability-weighted average of all possible outcomes, calculated by multiplying each outcome's value by its likelihood and summing the results to guide risk-based decision-making. ## Core Concept EMV answers: "If I face this decision repeatedly, what's my average gain or loss per trial?" It converts uncertain scenarios into a single expected value, enabling rational comparison between alternatives. **The Formula**: EMV = Σ(Probability × Impact) For a single outcome: EMV = P × I - P = Probability of occurrence (0-1 or 0-100%) - I = Financial impact (positive for opportunities, negative for threats) **The Innovation**: EMV transforms subjective uncertainty into quantified decision criteria. Instead of debating "should we?" it answers "what's the math?" ## When to Use **Ideal Scenarios**: - Project risk analysis with quantifiable financial impacts - Go/no-go investment decisions under uncertainty - Evaluating risk response strategies (mitigate vs. accept vs. transfer) - Comparing multiple projects with different risk profiles - Determining optimal bid pricing with win probability estimates - Resource allocation across uncertain opportunities **Not Suitable For**: - One-time, existential bets (EMV assumes repeatability) - Decisions with non-monetary values (reputation, lives, ethics) - When probabilities are purely guesswork (garbage in = garbage out) - Highly correlated risks (EMV assumes independence) - When risk aversion matters more than expected value ## Execution Steps ### 1. Identify All Possible Outcomes - List every distinct scenario that could occur - Include both positive outcomes (opportunities) and negative (threats) - Be exhaustive—missing outcomes invalidate the analysis - For complex decisions, use a decision tree to map branches *Output*: Complete list of mutually exclusive outcomes ### 2. Assign Probabilities - Estimate likelihood of each outcome (must sum to 100%) - Use historical data, expert judgment, or statistical models - Document assumptions (especially for subjective estimates) - Validate: Do probabilities reflect reality or wishful thinking? *Output*: Probability distribution across all outcomes ### 3. Quantify Financial Impact - Determine monetary value for each outcome - Positive values for gains (revenue, cost savings) - Negative values for losses (costs, penalties, rework) - Use consistent units (all in USD, EUR, etc.) - Include ALL financial effects (direct + indirect) *Output*: Impact values for each outcome ### 4. Calculate EMV for Each Outcome - Multiply probability × impact for each scenario - Example: 30% chance of $100K profit → EMV = 0.30 × $100K = $30K - Example: 10% chance of $50K cost → EMV = 0.10 × (-$50K) = -$5K - Keep signs correct (positive = gain, negative = loss) *Output*: Individual EMV per outcome ### 5. Sum to Total EMV - Add all individual EMVs (respecting positive/negative signs) - Total EMV = Σ(P₁×I₁ + P₂×I₂ + ... + Pₙ×Iₙ) - Positive EMV suggests favorable decision - Negative EMV suggests unfavorable decision *Output*: Single EMV figure for the decision ### 6. Compare Alternatives - Calculate EMV for each option (including "do nothing") - Rank by highest EMV - Consider risk tolerance (high EMV with high variance vs. low EMV with certainty) - Perform sensitivity analysis on key assumptions *Output*: EMV comparison table, recommended option ### 7. Document and Communicate - Show decision tree or calculation table - Explain assumptions behind probabilities and impacts - Highlight sensitivity to key variables - Recommend decision with rationale *Output*: Decision memo with EMV justification ## Common Pitfalls **"EMV is Gospel" Fallacy** EMV is a mathematical average, not a prediction of what WILL happen. You could face the worst-case scenario even if EMV is positive. **Solution**: Pair EMV with risk tolerance analysis. Know your "ruin threshold" (loss you cannot survive). **Probability Estimation Bias** People overestimate rare events (plane crashes) and underestimate common ones (project delays). Subjective probabilities are notoriously unreliable. **Solution**: Use reference class forecasting, historical data, or calibrated expert judgment (Tetlock's Superforecasting). **Ignoring Correlations** Calculating EMVs independently when risks are correlated (e.g., market crash affects all projects simultaneously). **Solution**: Use Monte Carlo simulation for correlated variables, or adjust probabilities for joint scenarios. **Overlooking Risk Aversion** EMV assumes risk neutrality. In reality, losing $1M hurts more than gaining $1M feels good. **Solution**: Apply utility theory (risk-adjusted EMV) or set EMV thresholds based on organizational risk appetite. ## Key Insights **PMP Exam Essential**: EMV is heavily tested on PMP certification. Common question format: "Given these probabilities and impacts, what's the EMV?" or "Which project has the highest EMV?" **Decision Trees = Visual EMV**: Decision trees represent EMV calculations graphically. Each branch shows probabilities, each end node shows impact, and EMV flows backward from outcomes to decision nodes. **EMV ≠ Most Likely Outcome**: A project with 90% chance of $10K profit and 10% chance of $200K loss has positive EMV (+$9K - $20K = -$11K)—yet the most likely outcome is a $10K profit. EMV captures expected value over many trials. **Opportunity vs. Threat**: EMV handles both. Positive impacts = opportunities to pursue; negative impacts = threats to mitigate. Combined EMV shows net expected position. ## Real-World Application **Project Bidding**: A contractor estimates 60% chance of winning a $500K profit bid, vs. 40% chance of losing $100K (preparation costs). EMV = (0.6 × $500K) + (0.4 × -$100K) = $300K - $40K = $260K → Bid is favorable. **Risk Response Strategy**: A project faces a 20% risk of $150K overrun. Mitigation costs $25K upfront but reduces risk to 5%. Option A (accept risk): EMV = -0.20 × $150K = -$30K. Option B (mitigate): EMV = -$25K + (-0.05 × $150K) = -$25K - $7.5K = -$32.5K → Accept risk is optimal. **Product Launch Decision**: Three scenarios: Best case (30%, +$2M), Base case (50%, +$500K), Worst case (20%, -$300K). EMV = (0.30 × $2M) + (0.50 × $500K) + (0.20 × -$300K) = $600K + $250K - $60K = $790K → Launch decision supported. ## Related Frameworks - **Decision Tree Analysis**: Visual representation of EMV calculations across sequential decisions - **Monte Carlo Simulation**: Advanced EMV calculation with probability distributions (not point estimates) - **Expected Utility Theory**: Risk-adjusted EMV accounting for diminishing marginal utility - **Value at Risk (VaR)**: Focuses on downside tail risk rather than expected average - **Real Options Analysis**: EMV applied to valuing strategic flexibility (option to delay/expand) - **Sensitivity Analysis**: Tests how EMV changes with varying probabilities/impacts ## Anti-Patterns **False Precision** Reporting EMV to the nearest dollar when probabilities are educated guesses. **Ignoring the Variance** Choosing a project with EMV of $100K ± $1M over one with EMV of $90K ± $10K without considering volatility. **One-and-Done Analysis** Calculating EMV once at project start and never updating as new information emerges. **Overcomplicating Simple Decisions** Spending 10 hours on EMV analysis for a $500 decision. Use heuristics for low-stakes choices. ## Score Justification **Framework Assessment**: 40/50 (Tier 1 - Canonical) - **Practitioner Weight (8/10)**: Standard practice in project management (PMBOK), finance, and strategy consulting. Used daily by PMs worldwide. Slight deduction: Often formulaic in practice rather than deeply understood. - **Clarity & Executability (9/10)**: Dead simple formula (P × I). Anyone with basic math can calculate EMV. Crystal clear steps. - **Proven ROI (7/10)**: Prevents irrational risk-taking and enables data-driven decisions. However, limited evidence of EMV directly causing superior outcomes (correlation ≠ causation). - **Novelty (6/10)**: Straightforward application of probability theory. Not particularly counter-intuitive, though the "average over many trials" framing trips up beginners. - **Cross-Domain Applicability (10/10)**: Universally applicable: project management, venture capital, product development, procurement, R&D portfolio optimization, disaster planning. **Notable**: PMBOK certification ensures millions of project managers know EMV. It's the foundation for more advanced frameworks like Real Options and Monte Carlo simulation.