Presentation is loading. Please wait.

Presentation is loading. Please wait.

Economics and Computation Week #11: Auction: applications

Similar presentations


Presentation on theme: "Economics and Computation Week #11: Auction: applications"— Presentation transcript:

1 Economics and Computation Week #11: Auction: applications
Pingzhong Tang 12/4/2018 Pingzhong Tang

2 Two applications of auctions
eBay (Taobao, Amazon, JD) Google ads (Bing, Yahoo!, Facebook) 12/4/2018 Pingzhong Tang

3 eBay 12/4/2018 Pingzhong Tang

4 eBay 12/4/2018 Pingzhong Tang

5 Gross transaction volume comparison (year of 2015)
Alibaba = Taobao+Tmall = =$426B Amazon = $ 88B eBay = $83B Walmart = $ 471B 2016: Alibaba surpasses Walmart 12/4/2018 Pingzhong Tang

6 Proxy bidding 12/4/2018 Pingzhong Tang

7 Rules and notations Taobao: Two bidders meet secret reserve! 12/4/2018
Pingzhong Tang

8 Example 12/4/2018 Pingzhong Tang

9 Recommendation by eBay
12/4/2018 Pingzhong Tang

10 A bidding war towards the end
12/4/2018 Pingzhong Tang

11 Potential downside of early bidding
12/4/2018 Pingzhong Tang

12 Searching for secret reserves
12/4/2018 Pingzhong Tang

13 eBay design choice 12/4/2018 Pingzhong Tang

14 Hard vs soft closing rules
12/4/2018 Pingzhong Tang

15 Taobao auction closing rules
1)司法拍卖出价延时的基本规则 在设置了出价延时的拍品竞拍结束的前2分钟(以系统接受竞价的时间显示为准),如果有竞买人出价竞拍,那么该次拍卖时间在此次出价的时间的基础上自动延时5分钟,循环往复直到没有竞买人出价竞拍时,拍卖结束。 2)司法拍卖出价延时举例说明 比如:假设某件拍品的拍卖结束时间是8月8日22点整,如果在8月8日21点58分15秒,有竞买人出价,那么系统将拍卖结束时间自动延长至8月8日22点03分15秒;如果在22点03分有竞买人出价,那么系统将自动延时到22点08分。 3)司法拍卖出价延时的效果 出价延时给拍卖用户带来的便利包括以下几点: 1.将给到竞买人更加充分的参与竞拍的机会。 2.避免因为网络或者电脑原因导致的延迟而错失出价机会。 12/4/2018 Pingzhong Tang

16 12/4/2018 Pingzhong Tang

17 Hard vs Soft closing rule
Pros and Cons Soft closing=Early bidding: shared information promotes common value Hard closing=late bidding: allow seller to choose a closing time that attracts large demand (cf ) 12/4/2018 Pingzhong Tang

18 12/4/2018 Pingzhong Tang

19 12/4/2018 Pingzhong Tang

20 12/4/2018 Pingzhong Tang

21 12/4/2018 Pingzhong Tang

22 12/4/2018 Pingzhong Tang

23 12/4/2018 Pingzhong Tang

24 12/4/2018 Pingzhong Tang

25 Reputation system design
Readings: Reputation and feedback systems in online platform markets. AER-2016. The limits of reputation in platform markets: an empirical analysis and field experiments. 12/4/2018 Pingzhong Tang

26 Google: sponsored search auctions
12/4/2018 Pingzhong Tang

27 12/4/2018 Pingzhong Tang

28 Statistics Google advertising revenue:
2011: $36,531M 2012: $43,686M 2013 first 3 quarters: $ 36,600 M Hal Varian: Google Chief Economist “What most people don’t realize is that all that money comes pennies at a time” References: Varian 2008: Position auctions Edelman et. al 2007: Internet advertising and generalized second price auctions 12/4/2018 Pingzhong Tang

29 12/4/2018 Pingzhong Tang

30 12/4/2018 Pingzhong Tang

31 12/4/2018 Pingzhong Tang

32 12/4/2018 Pingzhong Tang

33 12/4/2018 Pingzhong Tang

34 VCG for keywords auction
Same allocation rule as GSP Everyone pays the damages he causes by his participation 12/4/2018 Pingzhong Tang

35 12/4/2018 Pingzhong Tang

36 Model K positions k= 1, …, K N bidders i=1,…, N
Bidder i values position k at V= vixk vi is the value of a click to bidder i xk is the probability of a click at this slot Efficient allocation: v and x follow the same order 12/4/2018 Pingzhong Tang

37 12/4/2018 Pingzhong Tang

38 Observations of GSP Theorem. If bidders were to bid the same amount under VCG and GSP, then each bidder’s payment would be paygsp≥ payvcg Proof. Induction on positions. QED 12/4/2018 Pingzhong Tang

39 12/4/2018 Pingzhong Tang

40 12/4/2018 Pingzhong Tang

41 12/4/2018 Pingzhong Tang

42 12/4/2018 Pingzhong Tang

43 12/4/2018 Pingzhong Tang

44 12/4/2018 Pingzhong Tang

45 12/4/2018 Pingzhong Tang

46 12/4/2018 Pingzhong Tang

47 12/4/2018 Pingzhong Tang

48 12/4/2018 Pingzhong Tang

49 12/4/2018 Pingzhong Tang

50 12/4/2018 Pingzhong Tang

51 Acknowledgement eBay Slides adapted from David Parkes@Harvard
Google slides adapted from Jonathan 12/4/2018 Pingzhong Tang


Download ppt "Economics and Computation Week #11: Auction: applications"

Similar presentations


Ads by Google