Financial Management: Intro to Forecasts and Trends Flashcards Preview

BEC > Financial Management: Intro to Forecasts and Trends > Flashcards

Flashcards in Financial Management: Intro to Forecasts and Trends Deck (7):

Business Forecasting Defined:

Business Forecasting: The estimation of the value of a variable at some future point in time.

Forecasting Uses: Use of forecasting is common in various business circumstances, including:

  • Forecasting macro-economic factors - market growth, inflation rate, tax rate, etc.
  • Budgeting process - sales forecast, etc.
  • Forecasting demand for products - for inventory and production purposes
  • Forecasting for investment decisions - interest rates, commodity prices, currency exchange rates, etc.
  • And others



What are the two major types of Forecasting Approaches?

Business Forecasting Types:

  1. Qualitative Forecasting:
    • Based on judgment and opinion
    • Subjective - (based on experience)
    • Often bases on consensus
    • Useful when quantitative data is lacking
    • Useful for long-range forecasting
  2. Quantitative Forecasting:
    • Based on quantitative data and models
    • Objective - (based on observable phenomena)
    • Often based on mathematical calculations and determinations


Qualitative Methods:

Qualitative Methods:

  1. Executive Opinion -- Jury of executive opinion using collective judgment of executives and managers
  2. Market Research -- Employs customer or other surveys to determine belief, preferences, etc.
  3. Delphi Method -- Develops a consensus of an expert group using a multi-stage process to converge on a forecast. 


Quantitative Methods:

Quantitative Methods:

  1. Time Series Models -- Use patterns in past data to predict future values
    • Also called the Extrapolation Method
      • Extrapolation: To estimate (a value of a variable outside a known range) from values within a known range by assuming that the estimatedvalue follows logically from the known values.
    • The approach is not concerned with cause, just patterns in data.
  2. Causal Models -- Assume the variable being forecasted is related to other variables and makes projections based on assumptions.


Business Forecasting Time Horizons:

There are 3 time frames for forecasting purposes:

  • Short-term - From immediate future up to 3 months out
    • Time-series are most appropriate
  • Medium-term - From 3 mos to 2 years
    • Time series and causal methods are appropriate
  • Long-term - Periods longer than 2 years
    • Causal and qualitative methods are appropriate, especially Delphi. 


Business Forecasting Error:

Forecast Error: Measures how accurate a given forecast was for a prior forecast period.

  • It's measured as the difference between actual value and forecasted value
  • Smaller the difference = better the forecast

Three Common Forecast Error Measures:

  1. Mean Absolute Deviation (MAD): measures the average absolute values of forecast errors
  2. Mean Squared Error (MSE): measures the average sum of forecast errors squared
  3. Mean Average Percentage Error (MAPE): forecast error divided by actual value 



Which one of the following sets shows the most likely method appropriate for short-term and long-term forecasting?

 Short-term Forecasting    Long-term Forecasting  
 Time series models   Market research surveys 
 Causal models   Time series models 
 Time series models   Delphi method 
 Delphi method   Time series models 

 Short-term Forecasting    Long-term Forecasting  

 Time series models   Delphi method 

Times series models are most likely appropriate for short-term forecasting and the Delphi method is most likely appropriate for long-term forecasting. Time series models use past values or patterns to predict a future value or values, but the longer the forecasting period, the less likely will the past values or patterns be relevant to those future values. The Delphi method is a qualitative forecasting method that involves a group of experts developing a consensus using a multi-stage process to converge on a forecast, which is a particularly useful approach for long-term forecasting.