تاثیر سیاست های قیمت گذاری بر تنوع فروش در زمینه خرده فروشی سوپرمارکت

The bullwhip effect is recognized to be a significant cost driver in supply chains. One of the measures proposed frequently to counter bullwhip effect is price stability, through everyday low pricing (EDLP). However, this study suggests that with an auto-regressive (AR1) demand process, the use of constant, instead of dynamic pricing may result in lower profitability and higher-demand volatility. An optimal price and stocking level policy is developed with normally distributed demand in this study and the model is tested using parameters from a supermarket scanner data set to determine the impact of two pricing policies. The hypothesis that low prior-period demand leads to discounted pricing is also tested and partially supported.

There has been a long-standing belief among the operations management community that certain marketing behavior can adversely affect operational performance. For instance, excessive promotions and price discounting may drive up sales variability and increase cost levels, especially in the upstream stages of the supply chain through the bullwhip effect. This research looks into the issue of pricing behavior to compare one of the proposed solutions to this problem, everyday low pricing (EDLP), to the case of dynamic pricing, when demand can be characterized by an auto-regressive (AR1) process. A model is developed to optimize period price and inventory level for the retail level of the supply chain, drawing on extensions of the infinite period newsvendor problem. This model is tested using parameters extracted from actual scanner data in order to understand the implications of this model under realistic circumstances. Inventory control has been widely studied in the operations management literature, yet to date little attention has been given to the topic of retail sales variation resulting from inventory control procedures. This research extends the newsvendor model to consider the question of sales variation, which has important supply chain implications in reducing the bullwhip effect, the well-known amplification of variation as orders progress up the supply chain. Reduction in sales variation at the terminal end of the supply chain has the potential to improve the overall efficiency of the upstream suppliers. This paper is organized as follows. Section 2 reviews the relevant literature with respect to bullwhip effect in supply chains. Section 3 proposes a dynamic model which develops optimal inventory policy decisions when demand can be described as an AR(1) process. Section 4 describes a set of empirically based simulation experiments that are used to test the implications of the theoretical model using data and parameters from a US-based supermarket chain. Section 5 discusses conclusions, limitations, and suggests further research topics.

This research has shown that dynamic pricing may lead to higher profitability and reduced demand variability when demand is serially correlated. This result represents a useful addition to extant literature since there has been a major, implicit assumption in supply chain management literature that stable pricing will always reduce demand fluctuations, resulting in reduced “bullwhip” effects and lower costs in the supply chain. This analysis suggests that adopting constant pricing may actually increase variation when the impact of serial correlation is not considered. Constant pricing may also tend to reduce profitability of the retailer under the conditions of serial correlation. The magnitude of this reduction tended to be small, based on empirically observed parameters, but this reduction may be significant to retailers in highly competitive markets where firms operate with very thin profit margins, or where high levels of serial correlation are expected. Analysis of empirical pricing behavior yielded some evidence that the retailer reacted to prior periods in setting prices. This yields partial empirical support for the proposition that retail price-setting behavior can increase the magnitude of the bullwhip effect. There are several limitations to this study. The first limitation is that there are no menu costs in the model. Retailers may be legally required display price on each individual item, sometimes referred to as item pricing. Changing prices on items weekly, as this model suggests, would be costly for the retailer. These menu costs may absorb any of the profitability improvements suggested by the model. Advanced technologies, however, may mitigate menu costs under some circumstances, improving the utility of the approach suggested in this paper. Second, the analysis does not consider pricing behavior in context. This analysis assumed a monopolistically competitive market, where retailers do not take strategic actions based on behavior of competitors. It seems likely that there is a dimension of strategic consideration to pricing behavior. For example, retailers would be aware that a competitor is offering a discounted price on a particular product, and might be temped to respond to that discount. The role of ownership may also be important to pricing policy. This analysis assumed that pricing levels are set for the specific retail outlet. For retail chains, pricing may be set at a higher level for regional considerations. These retail chains may exhibit different pricing behavior than would an individual proprietor. Another limitation is that the model does not represent the pricing behavior as seen in the empirical data. Examine actual prices from the data set, the retailer appears to implement two different approaches to pricing. In the first approach, the retailer establishes a base price level with discounts from this price (bath soap, cereal, and shampoo.) The other approach appears to be a cycling between two price levels, with some discounting (beer.) Neither of these patterns closely fits the behavior of the optimal prices suggested by the model, although on average the actual prices were quite close to levels predicted by the model. It is also possible that actual pricing strategy is inducing consumer behavior as described in the literature review. Changing pricing strategy as recommended in this paper may affect the level of serial correlation in demand. The final limitation to note is that residuals to demand equation estimates are not precisely normally distributed. This suggests that there may be explanatory variables missing from the model. Demand parameters tended to be unstable on the sample products that we analyzed, which may cause difficulties in practice with applying pricing guidelines that we suggest. These limitations suggest several avenues for further research on this topic. First, the set of parameters should be explored more fully to better generalize this model. These parameters can then be compared to a more representative sample from scanner data sets, including different product categories, different retailers, different geographic regions, and the like to improve the generalizability of the model. A second improvement would be to explicitly consider menu costs in developing optimal price behavior with serial correlation. Also, the pricing behavior of different retailers should be studied. This paper suggests that dynamic pricing may reduce demand variability, holding cost (supplier pricing) constant. However, what is optimal behavior upstream in this case? From the supply chain perspective, considering supplier behavior in the context of this model is a logical next step. Dynamic pricing from the supplier may also tend to reduce volatility, but this behavior is likely to interact with the retailer's pricing behavior. Resolving this interaction may lead to a better understanding of how to reduce the bullwhip effect through pricing mechanisms in a supply chain. Thus this research will, we hope, lead to many new lines of inquiry, revisiting many issues that have been taken for granted in supply chain management literature.