Outline In-Class Experiment on First-Price Sealed-Bid Auctions Professor John Morgan: Internet Auctions The Winner’s Curse Hypothesis: Kagel and.

OutlineIn-Class Experiment on First-Price Sealed-Bid Auctions

Professor John Morgan: Internet Auctions

The Winner’s Curse Hypothesis: Kagel and Levin (1986)

Revenue equivalence and strategic equivalence hypotheses: Auction Formats

Dutch vs. English with Private Values Drawn from [0,1]

)(

)( ,0,0:*

***0

odd

oeeed

PPEp

PPEpppH

)(

)( ,1N

1-N :

*

***1

dd

eeed

PEP

PEP mPPH

1)2)(N(N

1)-(N 2

2)(N1)N(N

1)-(N :

22

2

2

VVH ed

Experimental Design

Hypothesis H0

Plots of Price Differences

Plots of Price Differences

Plots of Price Differences

Hypotheses H1 and H2

First vs. Second Price Sealed-Bid

)(

)( ,0,0:

2*2

1*1

*2

*1

'

0o

o

PPEp

PPEpppH

)( )( ,:

2*

2

1*

1*

2*

1'

1

PEPPEPmPPH

e

d'

V V , V , V Var P V V :Var PH

2

12211

2

Experimental Design

Hypothesis H’0

Test of Hypothesis H’0

Plot of Price Difference

Plot of Price Difference

Plot of Price Difference

Plot of Price Difference

Hypothesis H’1 and H’2

Over-bidding in First-Price Auction

Summary• Dutch versus English auctions

– Dutch oral auction yields prices <= optimal prices (depending on how valuations are assigned), while English auction prices not significantly > optimal price

– H1 and H2 supported for both auctions

• First-price versus Second-price sealed-bid auction– Second-price sealed-bid auction not significantly < optimal price, while

prices in the First-price sealed bid auctions significantly > optimal price– Similarly, H1’ and H2’ supported for Second-price sealed-bid auction but

not for First-price auction

• English and Second-price auctions are isomorphic in the latter rounds

• Behavior in all except first-price sealed bid consistent with RNNE