Electronic Markets and Adverse Selection

Electronic Markets and Adverse Selection

Adverse selection is based on an asymmetry of information between buyer and seller, where the seller has the knowledge advantage regarding a products' condition and value. Given this asymmetry or imbalance of information between buyer and seller, the seller is intent on getting the best price for mediocre or even poor merchandise, while the buyer must judge the product's quality based on a below-average price relative to perceived quality. When adverse selection exists in markets, the higher quality, higher priced products are driven out of the markets while low-priced products positioned only on information the buyer chooses to share crowd out higher priced models. According to the classic work of Akerlof (1970) and his article The Market for 'Lemons:' Quality Uncertainty and the Market Mechanism, the author states that given the asymmetry or imbalance of information between buyers and sellers, buyers will gravitate their prices towards the lemons. Car buyers focus on the differences in distance between lemon cars and high quality ones, and without perfect knowledge of the car's condition, drive prices down. Wolf and Muhanna (2005) have tested four hypotheses of the principles of adverse selection as they pertain to online auto auctions on Ebay Motors website. Their findings are covered in the remainder of this paper.

The following table provides an overview of their hypotheses, statistical models, conclusions and selections of independent and dependent variables.

Table derived from Wolf and Muhanna (2005) Ebay Study

Hypotheses

Adverse selection is present and more pronounced in online auctions.

Sellers will protect themselves from the effects of adverse selection by setting reserve prices for online auctions involving higher quality items.

Items with higher observable quality will be less likely to sell in online auctions.

Seller's reputation has a positive effect on the price buyers are willing to pay, ceteris paribus.

Sellers with higher reputations will be less like to set reserve prices.

Auctions conducted by sellers with higher reputations will more likely end in a sale.

Statistical Model structure of dependent and independent variables

Independent Variables: Car Year of Manufacture; Miles; high bid; unique negative ratings; unique positive ratings

Dependent Variables: Kelly Blue Book Price Data; Online Reputation Systems Rankings;

Correlation matrix points to statistical significance at the.01 level of confidence for High Bids being most influenced by Kelly Blue Book (.94), with inverse correlation to Age of the car (-. 590) and miles driven (-. 583).

Correlation of Seller's Rating (0.30) to Auction's Highest Bid correlates at the.01 level of confidence.

Conclusions

The conclusion seems incomplete as negative ratings of online sellers would need to strongly correlate with low bidding to make adverse selection completely proven in this study. The fact that there is no strong correlation in this study to negative rankings, a foundational element of asymmetric information, makes one see the research as incomplete in its design.

Selection of Independent and Dependent Variables

The research design is incomplete in that it does not capture sales lost due to prices being too high and statistically proving the full hypothesis of adverse selection. The selection of the studies' independent variables quantified and validates high trust emanating from proven reputation systems. It does not however show the full elasticity of adverse selection on the price bands of cars sold on Ebay Motors. There is no demand curve, and further, no demand variable captured in this analysis, making elasticity of demand impossible to measure. Without that, the full impact of adverse selection cannot fully be explained.

Refuting an Online Auction Adverse Selection Hypothesis