X hits on this document

PDF document

Searching for Google’s Value: - page 28 / 44





28 / 44

have incentives to set IPO prices closer to ensuing market values if the factors causing the low prices can

be circumvented. Underpricing results in lost IPO proceeds and greater dilution, and typically represents

the largest cost of the issue. On the other hand, if a degree of underpricing is optimal, companies have

every incentive to accurately set the IPO prices to achieve optimal underpricing given the level of the

stakes involved.

Our results contribute to IPO research in general in a variety of ways. The prediction market

results show that the information necessary to forecast the post-IPO price of Google’s stock existed in a

public forum and could be aggregated cheaply well in advance of the IPO. This provides evidence

against models where large payments to investors are required to overcome problems of asymmetric

information (whether issuers have more information about IPO values than investors as in Chemmanur,

1993, or investors have more information than issuers as in Benveniste and Spindt, 1989). This evidence

is independent of the unique features of the Google IPO and, hence, is generalizable. Because of the

restrictions that the auction mechanism put on allocations of shares, underpricing here is also inconsistent

with models that rely on pre-commitment to prices or underpricing and/or discretionary allocations of

shares (e.g., Benveniste and Spindt, 1989, Chemmanur, 1993, Loughran and Ritter, 2002) as a general

cause of IPO underpricing. The evidence is mixed on models in which underpricing addresses the

winner’s curse caused by asymmetric information across investors (as in Rock, 1986). Evidence from the

combination of the prediction markets and the unique features of the Google IPO is consistent with

symmetric information models when there is a future benefit to underpricing (e.g., Welch’s, 1989, model

of underpricing to drive bad firms from the market and achieve higher secondary offering prices; Booth

and Chua’s, 1996, future benefits of ownership dispersion increasing liquidity and decreasing the overall

cost of borrowing; or Tinic’s, 1988, and Hughes and Thakor’s, 1992, future benefits of reduced legal


On the practical side, there are a number of mechanisms that may help firms set IPO prices closer

to market values or set them closer to optimal underpricing. Here, we introduce the idea of using a

prediction market to do so. Our evidence suggests that such markets can be successful in forecasting


Document info
Document views133
Page views133
Page last viewedMon Jan 16 22:32:55 UTC 2017