Реверсивний аукціон: вибір торговельної стратегії

Abstract

Розглянуто питання вибору і обґрунтування торговельної стратегії під час реверсивного (зворотного) аукціону. Сформульовано гіпотезу щодо превалювання значущості сприятливих умов закупівлі над обраною торговельною стратегією. За допомогою методу агентного моделювання на основі програмного забезпечення AnyLogic створено модель зворотного аукціону, який функціонує за умовами системи комерційних закупівель Rialto. Результати моделювання показали, що умови виконання контракту, незважаючи на їх величезне значення для досягнення перемоги у конкурентній боротьбі, повинні застосовуватися у комплексі з розумною торговельною стратегієюThis paper is devoted to the issues of reasoning and selection the trading strategy for the competitive bidding during the reverse auction. This term refers to the real time online auction, where the sellers compete to obtain business from the buyer and prices typically decrease as the sellers underbid each other. This specific type of auction process is used in Ukraine within the public e-procurement system ProZorro for the government sector and open e-tendering system Rialto for the commercial procurement. Since e-procurement is the newest method of doing business in Ukraine, the development the trading strategies based on world experience and scientific research are more than actual and appropriate. To achieve this goal we selected the agent-based modeling method at the framework of AnyLogic software. The reverse auction’s model includes the price and non-price indicators and takes into account some parameters for each of the participants as the starting price, modified price (price, which considers the influence of non-price indicators, in this case – delay of payment and delivery period), minimum price (limit on how low price can be charged for the item), the chosen trading strategy. The model has number of assumptions, namely: participants always follow their strategies, and they know their competitors. The model assumes that the three firms participated at the reverse auction, and there were three trading strategies: a strong price decline at the first round, a gradual price reduction and a strong price decline at the final stage. Several rounds of experiments were conducted during the previous studies and on this basis we formulated a hypothesis about the prevalence of favorable purchase conditions above the chosen trading strategy. For further experiments we set the option for non-price indicators, their weight and conducted three rounds of the reverse auction. The most attractive non-price criteria had the participant № 2, the most attractive initial price parameters (the lowest starting price and modified prices) had the participant № 3. The winner of the 1st stage was the company № 3, because it had a strategy of strong price decline at the first round, so its bid and modified price considering the lowest starting values were the lowest among all participants. The winner of the 2nd stage also was the participant № 3, because it continued to reduce its bid (as opposed to participant № 2, which followed the strategy of strong price decline at the final round, and company № 1, which gradually reduced the price, but had a worst level of non-price indicators). The final stage of the reverse auction gave the victory to the participant № 2 through its strategy of strong price decline at the final round and the most attractive contract conditions. It is important that the absolute value of the company № 2 bid (excluding non-price indicators) was higher than participant № 3 bid, which proves the great importance of favorable contract conditions. The competition between the company № 2 and company № 3 was very aggressive because of participant № 3 continued to actively reduce its bids at the last two rounds of the auction. Moreover, the company № 3 was the winner of the first two rounds, that according to Ukrainian e-procurement systems’ rules provides an opportunity to make its bid for the next round after other participants, so it has the ability to adjust its decision considering the bids of other participants. Results of the modeling showed that the conditions of the contract (other words, nonprice indicators) despite of their enormous importance for the victory at the competition, should be used in close conjunction with the relevant trading strategy