Lesson 6 - Forecasting Flashcards
Forecasting
- all business organizations have a need for a process that produces forecasts of expected sales and expenses.
- forecasts need to be sufficiently accurate so company can anticipate cash flow needs and make cash related decisions.
- accurate forecasts are a matter of business survival
Which businesses are most at risk?
Those that are experiencing a high rate of growth and are experiencing growth in revenue and expenses.
Business runs a high risk of having a cash flow problem.
Cash Flow Problem
More money going out than is coming in, albeit in a short
Quantitative Forecasting Methods
Times Series Analysis 1. Moving Averages 2. Smoothing Methods 3. Trend Projections 4. Seasonal Influence 5. Others Multiple Regression Analysis
Qualitative Methods
- Bottom-up
- Scenarios
- Delphi
- Others
Forecast Model
A mathematical way of producing a forecast.
Moving Average Forecasts
Underlying idea is that actual sales figure for the next time period will not be very different from recent sales figures.
Uses most recent n data values as a basis for forecasting the next period data value.
Moving average based on…
… the fact that as a new observation in the time series is made, it replaces the oldest value in the time series, and a new average is calculated.
Forecast Errors
Actual data minus the forecasted data.
Mean Squared Error
aka MSE, use this to compare how well one forecasting method does vs. another.
Goal: Minimize the MSE
Method: summing all of the squares of the errors and then calculating average.
Weighted Moving Average
Assign weights the preceding data values when calculating moving averages.
Most recent data will get the highest weight, and that weight will decrease for preceding points.
*** only requirement is that the sum of the weights must equal 1.
Exponential Smoothing
Forecasting method that is similar to weighted moving averages.
Trend Projection and Seasonal Influence
Data that moves upward and downward, or in a seasonal up and down fashion.
Examples: Quarterly sales of bicycles for Company ABC
Trend #1: Sales over 16 qusrters show increase in sales every year
Trend #2: Sales are lowest in 2nd quarter and at a high point in 4th quarter in each year.
Fractional Forecasting
Simple, effective way to forecast groups of products.
Forecast total sales of various combined categories, use history to determine appropriate fractions to assign to categories - which will allow you to forecast sub-groups.
Example: Men’s summer suit sales. Total forecasted sales for this year are 400 suits. Last year 12% of those summer suit sales were double-breasted. You can forecast that you will sell close to 48 dbl-breasted suits this summer season.
Bottom-up Forecasts
- Method where (in contrast to top-down forecasting) the sales revenue estimates of each product or product line are combined to compute revenue estimate for the entire firm.
- generated in the trenches (sales force, departments)
- these individuals give insight about future demand based on their knowledge of customer history and industry