적응적 이월효과를 가진 동태적 가격결정 모형의 응용가능성

국문 초록 (Abstract)

현재의 가격이 미래의 수요에도 영향을 미치는 동태적인 가격효과가 존재할 때, 가격결정을 위해 동태적 가격결정 모형의 사용을 고려한다. 전통적인 동태적 가격결정 모형은 이월효과를 시간에 따라 항상 일정하게 가정하고 있다. 또한 이러한 가정을 완화하기 위해 확정적인 이월효과의 감소를 가정한 변동이월효과 모형도 있었다. 본 연구에서는 전통적인 동태적 가격결정 모형이 현재효과(Current-effect)와 전기(t-1) 판매량의 선형결합(linear combination)으로 구성되어 있어 전통적인 동태적 가격결정 모형은 지수평활법과 유사한 모형으로 이해할 수 있음을 보인다. 본 연구에서는 이월효과의 변동을 보다 일반화하기 위해 적응적 지수평활법으로 잘 알려진 Trigg & Leach(1967)를 이용하여 적응적 이월효과를 가진 동태적 가격결정 모형을 제안하고, 전통적인 동태적 가격결정 모형, 시간에 따라 이월효과가 확정적으로 감소하는 변동이월효과 모형과의 비교를 통해 적응적 이월효과를 가진 모형의 응용가능성을 검증하고자 한다. 제안된 모형은 한국 담배시장의 19개 브랜드별 판매량에 응용된다.

다국어 초록 (Multilingual Abstract)

When marketing activities have influence on future sales as well as present sales, we could consider the use of typical dynamic pricing model. This model assumes the constant carry-over effect. To release this assumption, there was the dynamic pricing model with the decreasing carry-over coefficient. In this paper, we show that the typical dynamic pricing model consists of the linear combination between the current-effect and the one period lagged sales. So, it could be seen as the exponential smoothing method. For the general carry-over effect, in this paper, we propose the dynamic pricing model with adaptive carry-over effect from the well-known adaptive exponential soothing model of Trigg & Leach(1967) and compare the proposed model with the typical dynamic pricing model and the dynamic pricing model with the decreasing carry-over effect. The proposed model is applied to the sales of 19 brands in Korean cigarette market. When y(t) is the sales, Pr(t) is the price, and u(t) is the disturbance term at time t, typical dynamic pricing model could be arranged as equation 1. For comparison, we estimated the dynamic pricing model with the decreasing carry-over effect as equation 2. We proposed the dynamic pricing model with adaptive carry-over effect from the well-known adaptive exponential soothing model of Trigg & Leach(1967) as equation 3. Estimation results of three alternative models are table 1 that it shows the best model for 19 brands named from A to S. Table 2 reports the brands which have the estimation problems in three alternative models. We chosen the best model with the following four criteria; 1) the parameter of absolute price, β should have negative value, 2) the parameter r in the decreasing carry-over effect model should have the value ranged from 0 to 1, 3) all estimated parameters should be statistically significant, and 4) the fitted value of best model (Adj-R2) has to have the highest value than the last two alternative models. In table 2, there are five decision reserved brands (C, J, K, O, R in italics) because they violate one and/or more criteria above listed. From the results, we could see an applicability of the proposed model with adaptive carry-over effect. Even if it performed better only four brands in total nineteen brands, it is necessary to analyze the price response on sales using the proposed model because there were the brands (L, P) that the proposed adaptive model was the only model having no estimation problem among alternative three models.

참고문헌 (Reference)

1 유필화, "가격정책론" 박영사 1998

2 Simon, Hermann, "Price Management" North- Holland 1989

3 Trigg, D. W, "Exponential Smoothing with an Adaptive Response Rate" 18 (18): 53-59, 1967

4 Koyck, L.M, "Distributed Lags and Investment Analysis" North-Holland Publishing Company 1954

5 Leeflang, Peter S. H, "Building Models for Marketing Decisions" Kluwer Academic Publishers 2000