what can be learned from classical inventory

12
What Can Be Learned from Classical Inventory Models? A Cross-Industry Exploratory Investigation (2007) Sergey Rumyantsev and Serguei Netessine

Upload: jasmina-tacheva

Post on 18-Dec-2015

215 views

Category:

Documents


2 download

DESCRIPTION

What Can Be Learned From Classical Inventory

TRANSCRIPT

What Can Be Learned from Classical Inventory Models? A Cross-Industry Exploratory Investigation (2007)

What Can Be Learned from Classical InventoryModels? A Cross-Industry Exploratory Investigation (2007)Sergey Rumyantsev and Serguei Netessine1. IntroductionClassical inventory models:- assume exogenous and usually static business environments, and rational agents -> normative in nature;derived at the product level and not the firm level (microscopic level -> firm as a black box);often do not account for many practical considerations;

Can insights from these models be used to explain the inventory dynamics of entire companies?1. IntroductionMacroeconomics:looks at firms outside the black box, by analyzing aggregate behavior;no inside perspective -> not useful to describe internal inventory drivers, e.g. lead times-inventory levels relationship.

Contribution 1: Check for consistency between the two perspectives, and show that classical inventory theory conclusions are useful.Contribution 2: Quantify the association between the environmental variables and inventories at the firm level inventory level elasticity.2. Literature ReviewFrom simple models (EOQ) to advanced inventory models that incorporate stochastic and correlated demands, multiple products, and multiple echelons.Pure operations research model is blind to data issues (Wagner, 2002).Malhorta et al. (2001), Chen et al. (2005, 2007), Gaur et al. (2005) provide empirical insights.Lieberman et al. (1999) analyze inventory dynamics in automotive industry.Hendricks et al. (2005), Cachon et al. (2007) investigate at the supply chain level.Newsvendor inventory model: Cohen et al. (2003), Olivares et al. (2004)3. Formulation of Research HypothesesReasons for inventory:- Production not at same time and place as demand (lead times); - rigid production capacity but variable demand; - economies of scale in handling inventories;- nonstationarity (seasonality, stochasticity) in demand and/or supply.3. Formulation of Research HypothesesTwo methodological challenges:what classical inventory theory insights will hold at aggregate data level across time and space structural properties will not, but monotone ones will;understand relationship between operational variables throughout operations and decision-making: beginning vs end of period.

Assumptions: rationality and stationary and independent demand (+ proxies for behavioral aspects)3. Formulation of Research HypothesesH1: Aggregate inventory level is positively associated with aggregate mean demand through a concave function. = combination of linear (newsvendor) and square-root (EOQ) functions.H2: Aggregate relative inventory level (i.e., the ratio of inventory to sales) is negatively associated with company size. = economies of scale.H3: Aggregate inventory level is positively associated with aggregate demand uncertainty. = buffer.H4: Aggregate inventory level is positively associated with aggregate procurement lead times. = buffer.H5: Aggregate inventory level is positively associated with aggregate product margins. = underage.H6: Aggregate inventory level is negatively associated with aggregate inventory holding costs. = overage.

4. Data Descriptionquarterly (~seasonal fluctuations) data containing 44 time points between 1992 and 2002 for 722 firms from the Compustat financial database.panel datapooled and segment-specific estimationsdata represents 30% of total US manufacturing and retailing business inventory, strongly correlated with total US inventory.financial accounting data crudely reflects actual processes within companies.data aggregated across product lines and production units. -> product variety in the future.5. Description of Variables

6. Model Specification