The flashcards below were created by user
enknjs
on FreezingBlue Flashcards.
-
Delphi technique
Expert based quantitative and qualitative forecasting. Most popular method.
-
Forecast Subject
What is being forecast
-
Forecast Horizon
length of forecast
-
Forecasting Techniques
Multiple Techniques, pick the best on for what you are forecasting
-
Simple Moving Average (SMA)
Calculate the average of revenues in the previous forecast periods to predict the next. It weighs all previous periods equally.
-
Exponential Smoothing (EXS)
Assigns different weights to the more recent data. Sum of the weights used needs to equal 100%.
-
Smoothing Constant or alpha
The weight applied to the data in the forecast.
-
Underforecast
Forecast is smaller than the actual number
-
Overforecast
The forecast is larger than the actual number
-
Transformation Moving Average (TMA)
Takes into account trends over time. In general it is more accurate than SMA and EXS over time
-
Revenue Trend
Trend occurs when the revenue shows a distinctive direction over time.
-
Upward Trend
The revenues have been getting larger overtime. Positive trend.
-
Downward Trend
The revenues have been getting smaller overtime. Negative trend.
-
Incremental changes
The difference in the numbers over time. Year 2- Year 1 = X; Year 3 - Year 2=Y; Average of X, Y, Z added to the last year of data.
-
Regression Against Time
A relationship is established between revenue and the forecast period. Creates a y=mx+b line where y is revenue and x is time. No evidence shows that it is more accurate than TMA for trend data.
-
Time-series forecasting
Forecasting techniques that use historical data. The historical data is called time-series data.
-
Quasi Casual Forecasting Model
Use when you have incomplete or skewed data. Use Revenue Predictors. Ex: when you know the size of your tax base and the tax rate you can estimate your tax revenue without having to base your forecast on historical data.
-
Absolute Percentage Error (APE)
Determines how accurate the forecast is. Absolute value (forecast value - actual value) divided by the actual value.
-
Mean Absolute Percentage Error (MAPE)
Determines how accurate the forecast is. Mean of Multiple APEs
-
Data Outlier
An obvious deviation from the mainstream revenue trend. Outliers lead to inaccurate forecast.
-
Steps in Revenue Forcasting
- 1. Cleaning the Data
- 2. Choosing the Forecast Technique (Forecast Accuracy and Cost)
- 3. Forecasting
- 4. Monitoring Forecast Performance
|
|