Ch.4 - Forecasting

technological forecasts

long-term forecasts concerned with the rates of technological progress - can result in the birth of exciting new products, requiring new plants and equipment.

the influence of forecasting on the product life cycle

- Introduction and growth require longer forecasts than maturity and decline - As product passes through life cycle, forecasts are useful in projecting: - staffing levels - inventory levels - factory capacity

short-range forecast

- a forecasting future time horizon - forecast that has a time span of UP TO one year - is generally less than 3 months - used for: - purchasing - job scheduling - workforce levels - job assignments - production levels - usually employs different methodologies than longer-term forecasting (focus on quantitative/mathematical model used more) - tend to be more accurate than longer-term forecasts

long-range forecast

- a forecasting future time horizon - generally 3 years or more in time span - used for: - new product planning - facility location research - capital expenditures - research and development - deal with more comprehensive issues and support management decisions regarding planning and products, plants and processes - broader, less quantitative methods are used more

medium-range forecast

- a forecasting future time horizon - generally spans from 3 months to 3 years - used for: - sales and production planning - budgeting - analysis of variation in production plans - deal with more comprehensive issues and support management decisions regarding planning and products, plants and processes - broader, less quantitative methods are used more

market survey

- a qualitative model of forecasting - A forecasting method that solicits input from customers or potential customers regarding future purchasing plans. - can help not only in preparing a forecast but also in improving product design and planning for new products - What consumers say, and what they actually do are often different - DOWNSIDE: forecasts can be overly optimistic based on customer input

sales force composite

- a qualitative model of forecasting - A forecasting technique based on salespersons' estimates of expected sales. - Estimates from individual salespersons are reviewed for reasonableness, then aggregated - forecasts are reviewed for realism, then combined at district and national levels to reach an overall forecast - Sales reps know customers' wants - DOWNSIDE: forecasts tend to be overly optimistic

jury of executive opinion

- a qualitative model of forecasting - A forecasting technique that uses the opinion of a small group of high-level managers to form a group estimate of demand. - sometimes augmented by statistical models - combines managerial experience with statistical models - relatively quick - however, group think becomes a disadvantage - lower relative efficiency

delphi method

- a qualitative model of forecasting - A forecasting technique using a group process that allows experts to make forecasts. - Panel of experts, queried iteratively - 3 types of participants 1. decision makers (consist of a group of 5 to 10 experts who will be making the actual forecast) 2. staff personnel (assist decision makers by preparing, distributing, collecting, and summarizing a series of questionnaires and survey results) 3. respondents (a group of people, often located in different places, whose judgments are valued, provides input to decision makers before forecast is made) - Iterative group process, continues until consensus is reached

seasonal/seasonality

- a time series component of demand - a data pattern that repeats itself after a period of days, weeks, months, or quarters - a regular pattern of up and down fluctuations in time series that tie to recurring events such as weather or holidays - changes due to weather, customs, holidays, seasonal periods (spring, summer, etc), and so on. - data pattern/changes occur within a single year - there are 6 common patterns - the presence of this makes adjustments in trend line forecasts necessary

random

- a time series component of demand - consists of erratic, unsystematic, "residual" fluctuations - due to random variations or unforeseen events - have a short duration and are nonrepeating - "blips" in data caused by chance and unusual situations - have no discernable patterns and cannot be predicted

trend

- a time series component of demand - is the gradual upward or downward movement of the data over time - changes due to things like income, population, age distribution, cultural views, etc. - typically has a duration of several years - straight line trending upward or downward

cyclical

- a time series component of demand - patterns in the data that occur every several years. - usually tied into the business cycle and are of major importance in short-term business analysis and planning. - repeating up and down movements - affected by BUSINESS CYCLE, political, and economic factors - has a duration of multiple years - is difficult to predict - This is because cycles include a wide variety of factors that cause the economy to go from recession to expansion to recession over a period of years.

naïve approach

- a time series quantitative model of forecasting - A forecasting technique that assumes that demand in the next period is equal to demand in the most recent period. - sometimes cost effective and efficient - can be a good starting point for comparison with more sophisticated models - simplest way to forecast - ex. if January sales were 68, then February sales will also be 68

exponential smoothing method

- a time series quantitative model of forecasting - A weighted-moving-average forecasting technique in which data points are weighted by an exponential function. - weights decline exponentially - most recent data is weighted the most - involves very little record keeping of past data and is fairly easy to use. - requires a smoothing constant (a) - ranges from 0 to 1 - subjectively chosen - is a weighting factor - when a trend is present, this must be modified

moving average method

- a time series quantitative model of forecasting - consists of a series of arithmetic means - USED WHEN THERE IS LITTLE OR NO TREND - used often for smoothing - provides an overall impression of data over time - uses a number of historical actual data values to generate a forecast for the next period's demand - are useful if we can assume that market demands will stay fairly steady over time. - effective in smoothing out sudden fluctuations in the demand pattern to provide stable estimates. - always lags behind actual values -> n = number of periods in the moving average

weighted moving average method

- a time series quantitative model of forecasting - is a part of the moving average method - USED WHEN SOME TREND OR PATTERN IS PRESENT - older data is usually less important - weights are assigned based on experience and intuition (are arbitrary) - weights can be/are used to place more emphasis on recent values - effective in smoothing out sudden fluctuations in the demand pattern to provide stable estimates. - lags behind actual data, but usually reacts more quickly to demand changes - This practice makes forecasting techniques more responsive to changes because more recent periods may be more heavily weighted. -> n = number of periods in the moving average)

quantitative forecasts

- a type of forecasting approach - Forecasts that employ mathematical modeling to forecast demand. - used when the situation is "stable" and historical data and/or associative variables exist - existing products - current technology - involves mathematical techniques

qualitative forecasts

- a type of forecasting approach - Forecasts that incorporate such factors as the decision maker's intuition, emotions, personal experiences, and value system. - used when the situation is vague and little data exist - new products - new technology - involves intuition and experience

linear regression

- an associative quantitative model of forecasting - means that this model usually considers several variables that are related to the quantity being predicted. - Once these related variables have been found, a statistical model is built and used to forecast the item of interest. - A straight-line mathematical model to describe the functional relationships between independent and dependent variables. - involves fitting a trend line to historical data points to project into the medium to long-range - Linear trends can be found using the least squares technique - least squares technique: -> (y^, or y hat) = a + bx -> y^ = value of dependent variable, or computed value of variable to be predicted -> a = y-axis intercept -> b = slope of regression line -> x = independent variable - the least squares method minimizes the sum of the squared errors (deviations) - look at ch.4 slides 64-70 to get equations for calculating least square method variables -> a = (y with line on top, or average of y) - b*(x line on top, or average of x) 1. find x 2. find y 3. find b 4. find a 5. y = a + bx

Mean absolute percent error (MAPE)

- forecast error measurement - The average of the absolute differences between the forecast and actual values, expressed as a percent of actual values. - ADVANTAGE: expresses error as a percent of the actual values, which are undistorted by a single large value - MSE and MAD have the disadvantage of having their values depend on the magnitude of the item being forecasted (means that forecasts can be very large numbers) 1. subtract your actual value from your forecasted value (finds your error) for specific period (make sure to take absolute value of this) 2. divide the error value by the actual value 3. multiply this value by 100 to get your absolute percent error 4. do steps 1-3 for each period 5. sum the % errors for each period 6. divide the summed % error value by the total number of periods - is perhaps the easiest value to interpret

Mean Squared Error (MSE)

- forecast error measurement - The average of the squared differences between the forecasted and observed values. 1. forecast errors are obtained by taking the actual value minus the forecasted value, and then squaring the difference (answer) 2. do this for each period (n), and then sum each error value for each period 3. square the sum of error values 4. divide the squared value by the total number of periods (n) - a low value of MSE is better because you want to minimize it - MSE exaggerates errors because it squares them - seems to accentuate large deviations due to the squared term - using this error measurement typically indicates that we prefer to have several smaller deviations rather than even one large deviation.

Mean absolute deviation (MAD)

- forecast error measurement - a measure of the overall forecast error for a model - This value is computed by taking the sum of the absolute values of the individual forecast errors (deviations) and dividing by the number of periods of data (n) - sum of absolute deviations = sum |deviations| / n 1. find your forecasted values with a given (a) value (exponential smoothing formula) 2. subtract actual values from forecasted values for each period 3. take the sum of the absolute deviations for each period, and then divide that value by the total number of periods - the lower the MAD value, the better (less error)

Realities of Forecasting

- forecasts are seldom perfect - outside factors we cannot predict or control often impact forecasts - most forecasting techniques assume an underlying stability in the system - product family and aggregated forecasts are more accurate than individual product forecasts

time series models of forecasting

- models of quantitative forecasting - A forecasting technique that uses a series of past data points to make a forecast. - predict on assumption that the future is a function of the past - look at what happened over a period of time and use a series of past data to make a forecast - based on a sequence of evenly spaced (weekly, monthly, quarterly, and so on) data points. - obtained by observing response variables at regular time intervals - forecasts based ONLY ON PAST VALUES, no other variables are important - assumes that factors influencing past and present will continue to influence in the future 1. naïve approach 2. moving averages 3. exponential smoothing 4. trend projection

The 5 quantitative models of forecasting

- time series models: 1. naïve approach 2. moving averages 3. exponential smoothing 4. trend projection - associative model: 5. linear regression

forecast error

-> forecast error = actual demand - forecast value -> forecast error = At - Ft -> Ft = forecast in period t -> At = actual demand in period t - Several measures are used in practice to calculate the overall forecast error. - These measures can be used to compare different forecasting models, as well as to monitor forecasts to ensure they are performing well. - 3 most popular measures: 1. Mean absolute deviation (MAD) 2. Mean squared error (MSE) 3. Mean absolute percentage error (MAPE)

The 6 common seasonal patterns

1. period = week; length = day; # of seasons = 7 2. period = month; length = week; # of seasons = 4-4.5 3. period = month; length = day; # of seasons = 28-31 4. period = year; length = quarter; # of seasons = 4 5. period = year; length = month; # of seasons = 12 6. period = year; length = week; # of seasons = 52

The 2 types of forecasting approaches

1. quantitative forecasts 2. qualitative forecasts

The 4 time series components of demand

1. trend 2. seasonal/seasonality 3. cyclical/cycles 4. random/random variations

The 7 steps in the forecasting system

1. Determine the use of the forecast 2. Select the items to be forecasted 3. Determine the time horizon of the forecast 4. Select the forecasting model(s) 5. Gather the data needed to make the forecast 6. Make the forecast 7. Validate and implement results

3 potential problems with moving average method

1. Increasing the size of n (the number of periods averaged) does smooth out fluctuations better, but it makes the method less sensitive to changes in the data. 2. Moving averages cannot pick up trends very well. Because they are averages, they will always stay within past levels and will not predict changes to either higher or lower levels. That is, they lag the actual values. 3. Moving averages require extensive records of past data.

Notes on least squares linear regression

1. We always plot the data because least-squares data assume a linear relationship. If a curve appears to be present, curvilinear analysis is probably needed. 2. We do not predict time periods far beyond our given database. For example, if we have 20 months' worth of average prices of Microsoft stock, we can forecast only 3 or 4 months into the future. 3. Deviations around the least-squares line are assumed to be random and normally distributed, with most observations close to the line and only a smaller number farther out.

3 types of forecasts

1. economic 2. technological 3. demand

The 4 parts of the product life cycle

1. introduction 2. growth 3. maturity 4. decline

The 4 qualitative models of forecasting

1. jury of executive opinion 2. delphi method 3. sales force composite 4. market survey

strategic importance of forecasting on capacity

Capacity shortages can result in undependable delivery, loss of customers, loss of market share

strategic importance of forecasting on supply-chain management

Good supplier relations and the ensuing advantages in product innovation, cost, and speed to market depend on accurate forecasts.

strategic importance of forecasting on human resources

Hiring, training, and laying off workers all depend on anticipated demand.

the forecast is the only estimate of demand UNTIL_________________

actual demand becomes known - forecasts therefore drive demands in many areas

good forecasts are an _________ part of efficient service and manufacturing operations

essential

a forecast is usually classified by the ___________ that is covers

future time horizon - 3 categories: 1. short range forecast 2. medium range forecast 3. long range forecast

economic forecasts

planning indicators that are valuable in helping organizations prepare medium to long range forecasts - address the business cycle by predicting inflation rates, money supplies, housing starts, and other planning indicators.

forecasting

process of predicting future events - the art and science of predicting future events - underlying basis of all business decisions: - production - inventory - personnel - facilities - may involve taking historical data (such as past sales) and projecting them into the future with a mathematical model. - may be a subjective or an intuitive prediction (e.g., "this is a great new product and will sell 20% more than the old one"). - may be based on demand-driven data, such as customer plans to purchase, and projecting them into the future.

demand forecasts

projections of a company's sales for each time period in the planning horizon - the focus is on rapidly identifying and tracking customer desires. - drive a company's production, capacity, and scheduling systems and serve as inputs to financial, marketing, and personnel planning.

the main purpose of forecasting

reduce uncertainty and make good estimates of what will happen in the future

forecasts tend to be more accurate as they become_______

shorter - therefore, forecast error also tends to drop with shorter forecasts

exponential smoothing with trend adjustment

simple exponential smoothing is like any other moving-average technique: It fails to respond to trends. - When a trend is present, exponential smoothing must be modified - The idea is to compute an exponentially smoothed average of the data and then adjust for positive or negative lag in trend. - With trend-adjusted exponential smoothing, estimates for both the average and the trend are smoothed. This procedure requires two smoothing constants: (a) for the average and (b) for the trend. 1. compute Ft 2. Compute Tt 3. calculate forecast: FITt = Ft + Tt - these steps and calculations are done for each period (month in book example) - The value of the trend-smoothing constant, (b), resembles the (a) constant because a high (b) is more responsive to recent changes in trend. - A low (b) gives less weight to the most recent trends and tends to smooth out the present trend. - Values of (b) can be found by the trial-and-error approach or by using sophisticated commercial forecasting software, with the MAD used as a measure of comparison.

smoothing constant (a)

the weighting factor used in an exponential smoothing forecast - a number greater than or equal to 0 and less than or equal to 1 - subjectively chosen - is generally in the range from .05 to .50 for business applications. - It can be changed to give more weight to recent data (when a is high) or more weight to past data (when a is low). - if the value is = 1, the forecast becomes identical to naïve model - Low values of a are used when the underlying average is fairly stable. - High values of a are chosen when the underlying average is likely to change.