بهینه سازی قیمت نقل و انتقالات برای مدیریت درآمد اتحاد خطوط هوایی مبتنی بر گزینه

Recently, an option-based capacity control process for the case of airline alliance revenue management with two partner airlines providing flight tickets on a single flight leg has been proposed. This previous work describes the determination of the booking limits as control variables for the capacity control by means of real options as well as simulation models which consider the option-based process to evaluate the booking process of the partner airlines within the strategic alliance. The booking limits are improved with simulation-based optimization in an iterative process. However, the transfer prices are assumed to be given. In this paper, the optimal transfer prices will be determined by a negotiation process. The results of the option-based capacity control process combined with transfer price optimization will be compared with the results of a first-come-first-served approach, ex post optimal solutions, a blocked seat allotment procedure and a random approach.

Due to deregulation of fares in the airline industry in the late 1970s, major airlines were confronted with the competition of low-cost carriers entering the markets. As stated by Shumsky (2006), low-cost competitors force major traditional carriers to process an increasing amount of their traffic in airline alliances. By forming strategic alliances, the airlines can generate additional revenues, for example, due to extended flight networks, coordinated flight schedules, and higher load factors. Further incentives for airlines to join strategic alliances are listed in Oum and Park (1997). According to O'Neal et al. (2007), the partner airlines within an alliance combine their flights through code sharing. Code-sharing agreements allow partner airlines within the alliance to offer a flight operated by one of the partners as a product of another partner airline. To maximize their profit generated from a limited seat capacity, the airlines decide which fares to charge and how many seats to reserve for each customer segment with support of revenue management instruments. Talluri and van Ryzin (2004) give a detailed description of revenue management instruments. Kimms and Klein (2005) list several specific and general definitions of revenue management and describe the revenue management instruments as well as the requirements for implementing revenue management instruments. In this paper, we focus on capacity control in revenue management applications in the airline industry. However, there are several other sectors in which the use of revenue management instruments make significant contributions to the performance. Kimms and Klein (2005) and McGill and van Ryzin (1999) present an overview of revenue management research in non-airline service sectors. Airlines use revenue management capacity control to coordinate the seat capacity of an aircraft. Talluri and van Ryzin (2004) outline current publications covering capacity control methods for a single airline not part of an alliance which in fact is already a highly complex problem. However, new decision problems concerning the capacity allocation occur if airlines build strategic alliances: Capacity control not only has to sort out how many seats should be allocated to the different booking classes of the airlines, but also how the seats will be divided among the alliance partners. Boyd (1998) specified two common decision control mechanisms used in practice: in a free sale, the airline operating the considered flight provides access to the seats in the aircraft to the non-operating alliance partners. The alliance partner airlines are allowed to access the seats, for example, in a first-come-first-served order. In a blocked seat allotment procedure, each airline will individually control the seats they have been assigned before the booking process. The drawbacks of capacity control methods so far applied for strategic alliances are: In a free sale setting, no capacity will be reserved for higher yielding booking classes while in a blocked seat allotment procedure, each airline will individually control the seats they have been assigned to which leads to static allotments. Although there are some publications regarding alliance revenue management (compare Boyd, 1998, Brueckner, 2003, Brueckner and Whalen, 2000, Vinod, 2005 and Wright et al., 2010), to the best of our knowledge, there is no literature that describes option-based capacity control models or methods with transfer price optimization for strategic alliances. We presented an option-based decision control for two partners within an alliance in a previous publication (see Graf and Kimms, 2011). This option-based decision control overcomes the mentioned disadvantages of the common decision control mechanisms so far used in practice by calculating booking limits for the alliance partners to reserve seat capacity for higher yielding booking classes and by allowing the alliance partners to switch their assigned capacity during the booking processes. The main contribution of our work to revenue management literature is an option-based capacity control procedure with transfer price optimization to divide the capacity among partners of a strategic alliance. In Graf and Kimms (2011), two procedures with the underlying option-based decision control were presented. The transfer prices used in these procedures were assumed to be given parameters. The surveys outlined in Graf and Kimms (2011) revealed that the results of the introduced methods depend on the choice of the transfer prices. The optimal transfer prices can be determined by systematically searching through the entire solution space. Since this approach is very run-time-intensive, methods to efficiently incorporate optimal transfer prices as an extension to the procedures described in Graf and Kimms (2011) will be introduced in this paper. This paper is organized as follows: In Section 2 we present the option-based capacity control procedure enhanced by a negotiation process to determine optimal transfer prices. Furthermore, we illustrate the determination of the booking limits and the simulation of the booking process of the alliance partners including real options and transfer prices. To be self-contained, 2.1 and 2.2 briefly repeat what is described in Graf and Kimms (2011) in greater detail already. In Section 3 the negotiation process to optimize the transfer prices will be discussed. Section 4 contains the computational study outlining the adopted test bed and comparing the results of the introduced option-based control with transfer price optimization to the results of the first-come-first-served approach, the ex post optimal solution, the blocked seat allotment approach, and the random approach. The implementation of the first-come-first-served approach, the ex post optimal solution, the blocked seat allotment approach and the random approach will be explained in Section 4. We summarize our study in Section 5 which concludes the paper with some comments on further research possibilities.

Revenue management capacity control in situations where multiple airlines build a strategic alliance is a difficult task due to the problem of distributing the seat capacity in an airline not only among the booking classes of an airline but also among the partner airlines. The OBP&SPSA approach introduced by Graf and Kimms (2011) describes a promising procedure for the distribution of the seat capacity among the booking classes of the partner airlines within an alliance. Unfortunately, the results of the proposed procedure highly depend on the choice of the transfer prices which are treated as given parameters. In order to improve the performance of the OBP&SPSA approach we presented an extension of the procedure (namely the option-based method with booking limit improvement through stochastic approximation and transfer price optimization) which optimizes the transfer prices in addition to the booking limit optimization. We studied the revenue management capacity control problem in the presence of two airlines building a strategic alliance in this paper. The non-operating partner within the alliance can reserve seats in a considered aircraft by means of real options. The main parts underlying the OBP&SPSA+Prices method were proposed: Deterministic linear option-based models were discussed to calculate the booking limits for the fare classes of the two airlines as well as simulation models which account for the option-based approach to simulate the booking processes of the airlines. Furthermore, the negotiation process between the two airlines was outlined describing the determination of the optimal transfer prices. In a computational study, the results of the OBP&SPSA+Prices method were compared to a first-come-first-served approach, ex post optimal solutions, a blocked seat allotment procedure, and a random approach. The survey shows, that the results of the OBP&SPSA+Prices approach are very promising overcoming the drawback of a first-come-first-served scenario (not reserving capacity for higher yielding classes and therefore accepting to many lower yielding requests) in all considered instances. Moreover, the results of the OBP&SPSA+Prices approach exceed the results of the blocked seat allotment procedure and the random approach in all considered instances. Considering the ex post optimal solutions, the determination of the optimal transfer prices improved the results of the OBP&SPSA+Prices approach toward the ex post optimal solutions. Since the run-time of the OBP&SPSA+Prices procedure is not critical, the presented method is applicable in revenue management systems of airlines. Future work may adjust the proposed procedure for the single-leg case to network revenue management problems, since many real-world problems appear in network revenue management settings. Moreover, the OBP&SPSA+Prices procedure should be generalized to allow the consideration of more than two partners within the alliance. Future research may also discuss the adoption of the OBP&SPSA+Prices procedure to solve capacity control problems in non-airline service sectors.