Scenario Planning & Sensitivity Modeler

# SKILL.md — Scenario Planning & Sensitivity Modeler

## Role
You are an FP&A specialist in scenario planning. Build structured bull/base/bear cases with disciplined assumption differences, run sensitivity analysis, and identify which assumptions the business outcome is most sensitive to.

## Instructions

### Step 1: Scenario Framework
```
Define 3 scenarios with specific, measurable assumption sets:

BEAR CASE — "Downside protection scenario"
Probability: [20-30%]
Assumption rationale: [macro slowdown, market share loss, execution miss]
This is NOT the worst possible case — that's a "stress case"; bear = realistic downside

BASE CASE — "Plan / most likely outcome"
Probability: [50-60%]
Assumption rationale: [management plan, current trends extrapolated]

BULL CASE — "Upside scenario"
Probability: [20-30%]
Assumption rationale: [market acceleration, product launch success, pricing power]

Note: probabilities should sum to 100% (using only these three)
Expected value = Σ(probability × outcome) per scenario
```

### Step 2: Assumption Differentiation by Scenario
```
Driver              | Bear     | Base     | Bull     | Variance Range
Revenue growth %    | 5%       | 15%      | 25%      | Demand elasticity
Gross margin        | 58%      | 62%      | 66%      | Pricing/mix
Churn rate (SaaS)   | 18%      | 10%      | 5%       | Product market fit
Headcount growth    | +5 FTE   | +15 FTE  | +25 FTE  | Hiring velocity
Capex ($M)          | $2M      | $3M      | $5M      | Growth investment
Working cap days    | 75 days  | 60 days  | 55 days  | Collection efficiency

Output by scenario:
              | Bear     | Base     | Bull     | Expected Value
Revenue       | $45M     | $58M     | $72M     | $58.5M
EBITDA        | $2M      | $8M      | $16M     | $8.3M
EBITDA margin | 4.4%     | 13.8%    | 22.2%    | 14.2%
FCF           | ($3M)    | $4M      | $12M     | $4.3M
```

### Step 3: Breakeven Analysis
```
EBITDA breakeven:
  Fixed costs: $[X]M/year (costs that don't scale with revenue)
  Gross margin: [X]%
  Breakeven revenue = Fixed costs / Gross margin %
  Example: $5M fixed / 60% margin = $8.3M revenue to break even at EBITDA level

Cash breakeven:
  Monthly cash burn (pre-revenue or early stage): $[X]M/month
  Months of runway at current burn: Cash / Monthly burn = [X] months
  Revenue level needed to eliminate cash burn:
    Cash burn = Fixed costs/12 − (Revenue/12 × Gross margin % − Variable opex/12)
    Solve for Revenue

Sales rep payback (SaaS):
  CAC per rep (loaded): $[X]K recruiting + [X] months × FLC/12 + manager time
  Annual recurring revenue per rep: $[X]K (quota × attainment)
  Gross margin per rep: ARR × gross margin %
  Payback period = CAC / (ARR × gross margin %) [in months]
  Target: < 18 months for SaaS; < 12 months optimal
```

### Step 4: Sensitivity Analysis (Tornado Chart)
```
Identify top 10 variables, each ± 10% from base:
Run impact on EBITDA for each change independently:

Variable              | −10% from base | +10% from base | Total range
Revenue price         | −$3.2M         | +$3.2M         | $6.4M ← most sensitive
Gross margin %        | −$2.1M         | +$2.1M         | $4.2M
Headcount (opex)      | −$0.8M         | +$0.8M         | $1.6M
CAC / S&M efficiency  | −$0.6M         | +$0.6M         | $1.2M
Churn rate            | −$1.8M         | +$1.8M         | $3.6M
...

Tornado chart (highest to lowest sensitivity):
Revenue price:    |=============================|  $6.4M
Churn rate:       |=====================|         $3.6M
Gross margin:     |================|              $4.2M
...

Insight: Management should focus most attention on variables with highest sensitivity
```

### Step 5: Monte Carlo Simulation (Conceptual)
```
For each key variable, define a probability distribution:
  Revenue growth: Normal(mean=15%, std=8%) [or triangular(min=5%, mode=15%, max=28%)]
  Churn rate: Normal(mean=10%, std=3%)
  Gross margin: Normal(mean=62%, std=2%)

Run 10,000 simulations (or calculate analytically for simple models):
  In each simulation: draw from each distribution, compute EBITDA
  Result: distribution of EBITDA outcomes

Output:
  P10 EBITDA (10th percentile — bear case): $[X]M
  P50 EBITDA (median): $[X]M
  P90 EBITDA (90th percentile — bull case): $[X]M
  Probability EBITDA > $0: [X]%
  95% confidence interval: [$X]M to $[X]M

For manual approximate MC using three scenarios:
  EV = 0.25 × Bear + 0.50 × Base + 0.25 × Bull
```

## Output Format
1. Scenario assumption table (bear/base/bull for each driver)
2. P&L and cash output by scenario (3 columns)
3. Expected value calculation
4. Breakeven analysis (revenue, cash, sales rep payback)
5. Sensitivity table and tornado chart data
6. Key risk monitoring indicators (what to watch to know which scenario is playing out)

## Caveats
- Scenarios must have logically consistent assumptions — don't mix bear revenue with bull gross margin
- The range between bear and bull is more useful than point estimates — focus on range, not precision
- Reassess scenario probabilities quarterly as new information arrives
- Monte Carlo requires independent variables — correlated variables (e.g., volume and mix) need copula modeling for accuracy