Post Stats

Costs

• sales results from stats -> vcos
• functions known
• derived from existing data
• period costs
• existing data
• assume constant
• managers will be able to change
• Problem
  This approach requires a working Financial Model to feed the default data and compute the variable cost formulas. Instead we should try to use Scenario Planning to encourage the spread of the Financial Model.

sales conversion

• why
• stats output = industry sales by size class by region
• desired cat sales (unit and $) by product model (invoices)
• issue
• what is a sale (shipment, end user sales)
• steps
• industry end user sales -> cat end user sales
  
  PINS model

  cat end user sales -> cat unit sales (from business unit perspective)

  Inventory/Distribution Model

  cat unit sales -> cat $ sales

  Pricing model
• PINS model
  • use PINS from WSA
  • PINS can be derived as constant or simple linear function
    Industry sales * PINS = cat sales
• Inventory/Distribution Model
  • if inventory levels constant
    End user Cat sales = Cat prodcution = Cat sales invoices
  • otherwise
    factory shipments = end user sales + delta dealer inventory
    factory production = factory shipments + delta factory inventory
    sales = some combination of factory shipments and end user sales
  • much work is involved
    to accurately figure out what "sales" are
    to feed inventory levels
• Pricing Model
  • need prices (from Marketing Model?)
  • need to know how much credit is given to each business unit for the sale
  • inflation - use current year dollars, ignore inflation, this won't affect ROA
  • how does marketing model tie in, might need marketing company help

• Jeff is starting work on connecting stat data derived by naive extraction to the financial model