financial-planning

name	financial-planning
description	FP&A financial planning — forecasting, scenario modeling, budgetary control.
tags	["finance","fpna","forecast","budget","scenario","planning","modelling"]

Financial Planning & Analysis (FP&A) Skill

When to Use This Skill

Use this skill when the user asks for:

• Forecasting (trend-based, seasonal, driver-based)
• Budget creation or variance-to-budget analysis -Scenario modeling and sensitivity analysis
• What-if analysis
• Capital expenditure (CapEx) planning
• Working capital projections
• Cash flow forecasting

Core Concepts

Forecasting Approaches

1. Trend-Based: Simple moving average or linear regression on historical data
2. Seasonal: Decompose into trend + seasonality components (e.g., Q4 spikes)
3. Driver-Based: Model driven by key metrics (headcount → payroll, revenue per employee → revenue)

Key FP&A Formulas

• Compound Annual Growth Rate (CAGR): (Ending Value / Beginning Value)^(1/Years) - 1
• Revenue per Employee: Total Revenue / FTE Count
• Operating Margin: (Operating Income / Revenue) × 100
• Working Capital: Current Assets - Current Liabilities
• DSO (Days Sales Outstanding): (Accounts Receivable / Revenue) × Days

Scenario Modeling Patterns

# Sensitivity Analysis Template
base_case = revenue * margin
scenarios = {
    "optimistic": revenue * 1.1 * margin,
    "pessimistic": revenue * 0.9 * (margin - 0.02),
    "best_case": revenue * 1.15 * (margin + 0.01),
}

Common SQL Patterns

Rolling 12-Month Trend

SELECT 
    period,
    SUM(revenue) OVER (ORDER BY period ROWS BETWEEN 11 PRECEDING AND CURRENT ROW) as rolling_12m
FROM sheet_
ORDER BY period;

Budget vs Actual by Category

SELECT 
    category,
    MAX(CASE WHEN type = 'budget' THEN amount END) as budget,
    MAX(CASE WHEN type = 'actual' THEN amount END) as actual,
    MAX(CASE WHEN type = 'actual' THEN amount END) - MAX(CASE WHEN type = 'budget' THEN amount END) as variance
FROM sheet_
GROUP BY category;

Best Practices

1. Document assumptions: Every forecast needs explicit assumptions documented
2. Scenario documentation: Note which scenario base case vs upside/downside
3. Bridge building: Build clear bridges from forecast to prior period
4. Sensible rounding: Present millions/thousands, not decimals
5. Include context: Add narrative to explain "why this number"

Avoiding Common Mistakes

• Don't forecast without first Understanding data quality
• Don't ignore one-time events that skew historicals
• Don't overcomplicate models (simpler is often better)
• Don't forget to align totals across statements (PL, BS, CF)
• Don't present without context (what does the number mean?)