A-CAT CORP.: FORECASTING Dr. Ravi Kumar, PhD NTPC School of Business, Noida Brief Introduction of Case • A-CAT Corp. produced electrical appliances in India and largely catered to the price-sensitive rural market. • These customers are quite sensitive about pricing. • In the opinion of top level management, there was more scope in this segment. • Recently, production department has complaining shortages of spare and components, especially the major ones. • The vice-president asked operations manager (Shirish Ratnaparkhi) to focus on the issue and come up with some practical solution ( forecasting report). • The operations manager’s job was to collect data, analyse data patterns, use forecasting methods, and carry out back testing. Issue • In recent months, the sales of voltage regulators at A-CAT had shown a very disturbing trend. • In reaction, A-CAT wanted to look at their policy of purchasing and stocking spare and components in the system, with regards to schedule and stock-inhand inventory. • The brief was to prepare a comprehensive plan for better transformer purchasing, using various forecasting techniques, and thoroughly analyse results and finding. Earlier, A-CAT has four suppliers, now only one. Forecasting • Forecasting is the art and science of predicting future. • Lead times require that decisions be made in advance of uncertain events. • Forecasting is important for all strategic and planning decisions in a business organization. • Forecasts of product demand, materials, labor, and financing are an important inputs to scheduling, acquiring resources, and determining resource requirements. Demand Management • Demand management is the interface between manufacturing planning & control and marketplace. • Activities include Forecasting Order Processing Making Delivery Promises Forecasting Horizons • Short Term (0 to 3 months): for inventory management and scheduling. • Medium Term (3 months to 2 years): for production planning, purchasing, and distribution. • Long Term (2 years and more): for capacity planning, facility location, and strategic planning. Principles of Forecasting • Forecasts are never accurate. • Every forecast must include an estimate of error. • The greater the degree of aggregation, the more accurate the forecast. • Long-term forecasts are usually less accurate than the shore-term forecasts. Factors Affecting Forecasting Methods • Time Frame • Demand Behaviour • Causes of Behaviour Demand Behaviour • Trend a gradual long-term up or down movement of demand • Cycle an up and down repetitive movement in demand • Seasonal Pattern an up and down repetitive movement of demand occurring periodically • Random Variations Movements in demand that do not follow a pattern Demand Demand Random movement Time (b) Cycle Demand Demand Time (a) Trend Time (c) Seasonal pattern Time (d) Trend with seasonal pattern Forecasting Methods • Qualitative Use management judgement, expertise, and opinion to predict the future demand • Quantitative Time Series: Statistical techniques that use historical demand data to predict the future Regression: Attempts to develop a mathematical relationship between demand and factors that cause its behaviour Forecasting Process 1. Identify the purpose of forecast 2. Collect historical data 3. Plot data and identify patterns 6. Check forecast accuracy with one or more measures 5. Develop/compute forecast for period of historical data 4. Select a forecast model that seems appropriate for data 7. Is accuracy of forecast acceptable? 8a. Forecast over planning horizon 8b. Select new forecast model or adjust parameters of existing model 9. Adjust forecast based on additional qualitative information and insight 10. Monitor results and measure forecast accuracy Time Series • Assume that what has occurred in the past, will continue to occur in the future. • Relate the forecast to only one factor – time. • Include: • Moving Average • Exponential Smoothing • Linear Trend Line Moving Average • Naïve Forecast • Demand of current period is used as next period’s forecast • Simple Moving Average • Stable demand with no pronounced behavioural pattern • Weighted Moving Average • Weights are assigned to most recent data Simple Moving Average n D i=1 i MAn = n where n = number of periods in the moving average Di = demand in period i Weighted Moving Average WMAn = Wi Di i=1 where Wi = the weight for period i, between 0 and 100 percent W = 1.00 i Exponential Smoothing • Averaging method • Weights most recent data more strongly • Reacts more to recent changes • Widely used, accurate method Ft +1 = Dt + (1 - )Ft where: Ft +1 = forecast for next period Dt = actual demand for present period Ft = previously determined forecast for present period = weighting factor, smoothing constant Adjusted Exponential Smoothing AFt +1 = Ft +1 + Tt +1 where T = an exponentially smoothed trend factor Tt +1 = (Ft +1 - Ft) + (1 - ) Tt where Tt = the last period trend factor = a smoothing constant for trend Linear Trend Line y = a + bx where a = intercept b = slope of the line x = time period y = forecast for demand for period x xy - nxy b = x2 - nx2 a = y-bx where n = number of periods x x = = mean of the x values n y y = n = mean of the y values Seasonal Adjustments • Repetitive increase/decrease in demand • Use seasonal factor to adjust the demand Seasonal factor = Si = Di D Forecast Accuracy • Forecast Error • MAD (Mean Absolute Deviation) • MAPD (Mean Absolute Percent Deviation) • RSFE (Running Sum of Forecast Error or Cumulative Error) • Average Error or Bias MAD Dt - Ft MAD = n where t Dt Ft n = period number = demand in period t = forecast for period t = total number of periods = absolute value Mean absolute percent deviation (MAPD) |Dt - Ft| MAPD = Dt Cumulative error E = et Average error et E= n |Actual - Forecast| Mean Absolute MAPE = Percent Error Actual n • Tracking signal • monitors the forecast to see if it is biased high or low Tracking signal = • 1 MAD ≈ 0.8 б (Dt - Ft) E = MAD MAD • The “Tracking Signal” quantifies “Bias” in a forecast. No product can be planned from a badly biased forecast. Tracking Signal is the gateway test for evaluating forecast accuracy. • Tracking signal is computed as the running sum of forecast error (RSFE) divided by MAE. We compute RSFE by summing up the forecast errors over time.