【作者】 杨琴；

【导师】 王宗军；

【作者基本信息】 华中科技大学 ， 工商管理， 2018， 博士

【摘要】 随着互联网的蓬勃发展,全球搜索引擎用户规模不断增加。搜索引擎利用关键词拍卖向搜索引擎用户定向提供信息的同时,通过推广商业广告获取巨大盈利,并给广告主带来高额回报,这是传统拍卖交易理论在付费搜索领域的创新性运用。在搜索引擎注重质量权重拍卖规则的推动下和广告主自身利益的驱使下,广告主会通过不断投资来提高自身表现水平,从而吸引更多用户点击,研究广告主投资行为及其产生的影响意义深远。基于此,本文从广告主内生投资的角度入手,考虑参与者风险规避特性以及搜索引擎保留价下,综合运用博弈论、概率论、比较静态分析、最优化理论和数值分析等方法,探讨广告主投资行为和竞价策略以及搜索引擎满足收益最大化和社会福利最大化的最优保留价,以期扩展关键词拍卖理论在广告主投资、风险规避以及搜索引擎关于保留价最优机制设计领域的研究。论文的主要研究工作如下:首先,将内生连续变化的投资刻画到广告主最终能实现的点击量里,根据初始表现水平将广告主分为高低两种类型,构建广告主投资竞价两阶段模型,分析两类广告主的最优投资决策和均衡竞价策略。基于广告主均衡竞价策略,着重研究决定低类型广告主类型转换的估价阈值的存在性和唯一性,分情况讨论了两类广告主的最优投资量,并进行对比分析。对搜索引擎拍卖规则对低类型广告主类型转换以及两类广告主最优投资量的影响进行了探讨,并通过数值分析进行验证。其次,在传统拍卖理论中,有关风险态度不同的竞标人理论和实证研究比较充分,而涉及搜索引擎关键词拍卖理论方面,扩展风险中性为风险规避情形的研究亟待发展。针对关键词拍卖中广告主不满足风险中性假设的情形下,引入风险规避型广告主。在广告主内生投资和搜索引擎设置保留价的拍卖规则下,探讨风险规避型广告主的均衡竞价策略和最优投资决策,以及广告主风险规避程度、广告主数量和搜索引擎保留价对广告主竞价和投资的影响。同时,通过数理推导得出搜索引擎最大化自身收益和最大化社会福利保留价的满足条件和搜索引擎收益的变化趋势。针对广告主风险规避程度和数量对搜索引擎保留价策略和收益的影响,采用数值算例进行了扩展研究。同时,进一步对比分析了保留价对搜索引擎收益和社会福利影响。最后,放宽关键词拍卖机制中搜索引擎是风险中性的假设,假定搜索引擎为风险规避,这种假设条件的释放更能体现出搜索引擎偏好的多样性。在风险规避型搜索引擎设置保留价的拍卖规则下,探讨风险中性广告主的均衡竞价策略和最优投资决策,并通过数理推导和算例验证分析广告主数量和搜索引擎保留价对广告主竞价及投资决策的影响。同时,关于广告主数量和搜索引擎风险规避程度对搜索引擎最优保留价策略和收益的影响以及搜索引擎如何设定保留价来权衡自身收益和社会福利,文中运用数值算例展开了探讨。更多还原

【Abstract】 With the flourishing development of the Internet,the scale of global search engine users has been increasing.Search engine uses keywords auction to provide information to Internet search engine users and promotes commercial advertising in order to gain high profits.This brings high returns to advertisers.It is an innovative use of the traditional auction transaction theory in the field of keywords auctions.Driven by the search engine’s auction rules which have more and more attention to the weight of the advertisers and advertisers’ own interests,the advertisers invest to improve their performance in order to attract more users to click.Studying advertisers’ investment behavior and its impact has far-reaching implications.Based on this background,this dissertation introduces the endogenous investment of advertisers,considering the participants’ risk aversion characteristics and search engine’s reserve price.Multiple analysis methods including game theory,probability theory,comparative static analysis,optimization theory and numerical analysis,etc.are adopted to explore the bidding strategies and investment decision of advertisers,as well as the optimal reserve price that maximizes the revenue of search engine or the social welfare.The analysis above is intented to expand the research on keywords auction theory in the fields of advertiser investment,risk aversion and search engine’s reserve price setting.The main research contents are generalized as follows:Firstly,we introduce the endogenous investment into the clicks which advertisers can finally achieve and divide advertisers into two types according to their initial performance.A two-stage model is established which consists of investment stage and bidding stage in order to analyse the two types of advertisers’ optimal investment decisions and equilbrium bidding strategy.Based on the equilibrium bidding strategy of advertisers,this paper focuses on the existence and uniqueness of the cut-off valuation for low-type advertisers in type conversion,and then compares the optimal investment of the two types of advertisers under different situation.In addition,it explores the influence of search engine auction rules on low-type advertisers’ type conversion and the optimal investment of advertisers.These conclusions are verified by numerical analysis.Secondly,in traditional auction theory,there are quite adequate resrarch on bidders with different attitudes toward risk from both the theoretical and empirical aspect.However,the research on extending the risk-neutral model to risk aversion in keyword auction theory needs prompt development.Considering the situation that the risk-neutral assumption of the advertiser under the keyword auction is not necessarily satisfied,risk-averse advertisers are introduced.This study which consists of advertisers’ endogenous investment and search engine’s reserve price explores the equilibrium bidding strategy and the optimal investment decision-making of the risk-averse advertisers.Then,sensitivity analysis on bidding and investment about the degree of advertisers’ risk aversion and the number of advertisers,as well as the reserve price is thoroughly discussed.Meanwhile,this dissertation deduces the determinant solution of the search engine’s optimal reserve price under the goals of maximizing search engine revenue and maximizing social welfare.Meanwhile,the tendency of search engine’s revenue is derived theoretically through mathematical method.Using numerical examples,we analyze the impact of the risk aversion of advertisers and the number of advertisers on search engine’s optimal reserve price and revenue.At the same time,the impact of reserve price on search engine’s revenue and social welfare is further analyzed.Finally,this paper relaxes the assumption that search engine is risk-neutral.This release of such assumptions can better reflect the diversity of search engine preferences.Under the auction rules that the risk-averse search engine set the reserve price,we discuss the equilibrium bidding strategy and the optimal investment decision of risk neutral advertisers.This paper analyzes the influence of the number of advertisers and search engine’s reserve price on the bidding strategies and investment decisions of advertisers with mathematical discussing and numerical examples.At the same time,how the number of advertisers and search engine’s risk aversion influence the search engine’s optimal reserve price and revenue,as well as how search engine sets the reserve price rules to balance its own revenue and social welfare are discussed by numerical examples.更多还原