【正文】 isk Pooling.l High coefficient of variation, which increases required safety stock.l Negatively correlated demand. Why?162。 The benefits from risk Pooling depend on the behavior of demand from one market relative to demand from another.Risk Pooling:Types of Risk Pooling*162。 Risk Pooling Across Markets162。 Risk Pooling Across Products162。 Risk Pooling Across Timel Daily order up to quantity is:? LT?AVG + z ? AVG ? ?LT10 1211 13 14 15DemandsOrders To Centralize or not to Centralize162。 What are the tradeoffs that we need to consider in paring centralized distribution systems with decentralized distribution systems?162。 What is the effect on:l Safety stock?l Service level?l Overhead costs?l Customer Lead time?l Transportation Costs?Centralized vs Decentralized system162。 Safety stock. Clearly, safety stock decreases as a firm moves from a decentralized to a centralized system.162。 Service level. When the centralized and decentralized systems have the same total safety stock, the service level provided by the centralized system is higher.162。 Lead time. Since the warehouses are much closer to the customers in a decentralized system,response time is much lower.162。 The warehouse echelon inventorySupplierWarehouseRetailers Managing Inventory in the SC (Centralized Systems*)The warehouse policy controls its echelon inventory position, that is, whenever the echelon inventory position for the W is below s, an order is placed to raise its echelon inventory position to S.Centralized Distribution Systems*162。 Question: How much inventory should management keep at each location?162。 A good strategy:l The retailer raises inventory to level Sr each periodl The supplier raises the sum of inventory in the retailer and supplier warehouses and in transit to Ssl If there is not enough inventory in the warehouse to meet all demands from retailers, it is allocated so that the service level at each of the retailers will be equal. Inventory Management: Best Practice162。 Periodic inventory reviews162。 Tight management of usage rates, lead times and safety stock162。 Introduce or enhance cycle counting practice162。 ABC approach162。 Reduced safety stock levels 162。 Shift more inventory or inventory ownership to suppliers 162。 Quantitative approachesChanges In Inventory Turnover162。 Inventory turnover ratio = annual sales/avg. inventory level162。 Inventory turns increased by 30% from 1995 to 1998162。 Inventory turns increased by 27% from 1998 to 2023162。 Overall the increase is from turns per year to over 13 per year over a five year period ending in year 2023.Inventory Turnover RatioFactors that Drive Reduction in Inventory 162。 Top management emphasis on inventory reduction (19%)162。 Reduce the Number of SKUs in the warehouse (10%)162。 Improved forecasting (7%)162。 Use of sophisticated inventory management software (6%)162。 Coordination among supply chain members (6%)162。 OthersFactors that Drive Inventory Turns Increase162。 Better software for inventory management (%)162。 Reduced lead time (15%)162。 Improved forecasting (%)162。 Application of SCM principals (%)162。 More attention to inventory management (%)162。 Reduction in SKU (%)162。 OthersForecasting 162。 Recall the three rules162。 Nevertheless, forecast is critical162。 General Overview:l Judgment methodsl Market research methodsl Time Series methodsl Causal methodsJudgment Methods162。 Assemble the opinion of experts162。 Salesforce posite bines salespeople’s estimates162。 Panels of experts – internal, external, both162。 Delphi methodl Each member surveyedl Opinions are piledl Each member is given the opportunity to change his opinionMarket Research Methods162。 Particularly valuable for developing forecasts of newly introduced products162。 Market testingl Focus groups assembled.l Responses tested.l Extrapolations to rest of market made.162。 Market surveysl Data gathered from potential customersl Interviews, phonesurveys, written surveys, etc.Time Series Methods162。 Past data is used to estimate future data162。 Examples includel Moving averages – average of some previous demand points.l Exponential Smoothing – more recent points receive more weightl Methods for data with trends:? Regression analysis – fits line to data? Holt’s method – bines exponential smoothing concepts with the ability to follow a trendl Methods for data with seasonality? Seasonal deposition methods (seasonal patterns removed)? Winter’s method: advanced approach based on exponential smoothingl Complex methods (not clear that these work better)Causal Methods162。 Forecasts are generated based on data other than the data being predicted162。 Examples include:l Inflation ratesl GNPl Unemployment ratesl Weatherl Sales of other productsSelecting the Appropriate Approach:162。 What is the purpose of the forecast?l Gross or detailed estimates?162。 What are the dynamics of the system being forecast?l Is it sensitive to economic data?l Is it seasonal? Trending?162。 How important is the past in estimating the future?162。 Different approaches may be appropriate for different stages of the product lifecycle:l Testing and intro: market research methods, judgment methodsl Rapid growth: time series methodsl Mature: time series, causal methods (particularly for longrange planning)162。 It is typically effective to bine approaches.演講完畢，謝謝觀看！