Print
E-MailJay R. Ritter is the Cordell Eminent Scholar at the University of Florida’s Warrington College of Business. This post is based on a recent paper authored by Professor Ritter; Chunxin Jia, professor at the Guanghua School of Management at Peking University; Zhen Xie, Bank of Suzhou; and Donghang Zhang, associate professor at the University of South Carolina’s Darla Moore School of Business.
From 2009-2012, China’s initial public offering (IPO) market was characterized by a regulatory environment in which the government did not control offer prices. If there was excess demand, underwriters were required to allocate shares on a pro rata basis, without favoring one group of clients over another, in contrast to the bookbuilding procedure that is widely used in the West. Furthermore, coverage by a stock before it went public by analysts working for brokerage firms not involved in the underwriting of an IPO was widespread, with an average IPO having coverage from 10 unaffiliated analysts before trading starts.
During our 2009-2012 sample period, underpricing (first-day returns) on Chinese A-share IPOs averaged 35%. Following a government-imposed moratorium during 2013 in which there were no IPOs, during 2014-2016 underpricing has averaged 219%. The recent severe underpricing is due to a government policy of limiting the offer price to a price-earnings ratio of no more than 23.
We use the institutional environment during 2009-2012 to examine the effect of pre-IPO analyst coverage and analyst optimism, measured by earnings forecasts, on the pricing of IPOs. We find that when there are more analysts and when the analysts are more optimistic in their earnings forecasts, the offer price is revised upwards by more, allowing the firm to raise more money. We also find that larger increases in the offer price are associated with less underpricing. This pattern is unlike the positive relation, known as the partial adjustment phenomenon, that exists in the U.S. and Europe when bookbuilding is used. We attribute the lack of partial adjustment to the inability of underwriters in China to allocate shares on a discretionary basis. Without this discretionary ability, there is an alignment of incentives between underwriters and issuers in China. With bookbuilding, underwriters have the ability to allocate shares to their favored clients, who might include hedge funds or others that are highly profitable clients. Hedge funds and mutual funds are permitted to pay extra commissions (known as soft dollars) on other transactions to underwriters in return for receiving favorable allocations of underpriced IPOs. This rent-seeking behavior creates an incentive for underwriters to not boost the offer price as much as they could when there is strong demand, resulting in the partial adjustment phenomenon. The full adjustment of offer prices to strong demand in China is also consistent with the mechanism design theory of bookbuilding, which posits that partial adjustment occurs when underwriters have discretion to allocate shares, because they need to allocate underpriced shares to institutional clients that provide useful information for pricing the offering.
In the U.S., affiliated analysts do not publish research reports until the end of the quiet period. Although the quiet period restrictions were lifted or relaxed in 2012 and 2015, industry practice continues to be that unaffiliated analysts generally do not initiate coverage of an IPO, and affiliated analysts wait 25 calendar days before initiating coverage. From a regulation perspective, there are benefits and concerns regarding pre-IPO communications from analysts. For investors, more pre-IPO communications such as pre-IPO analyst research help reduce information asymmetry and improve transparency. On the other hand, pre-IPO communications can also be used to condition the market and hype an IPO, increasing the demand.
Analysts working for the underwriters have an incentive to hype an IPO in order to achieve a higher offer price and/or leave more money on the table if the underwriters have allocation discretion. But all analysts, including unaffiliated analysts, have incentives to be excessively optimistic for three other reasons. First, if issuing companies prefer to hire investment banking firms for future deals that provide positive coverage, analysts have an incentive to cover a company and make optimistic forecasts. Second, if management is unwilling to talk to analysts that have a negative recommendation on the firm, a pessimistic analyst is put at an informational disadvantage. Third, owners of a stock want an analyst to be publicly optimistic. This latter incentive may have been especially strong in China during our sample period because short-selling was prohibited until 2010, although it is still prohibited for IPOs during their first three months of trading.
We have some evidence that potential agency concerns exist for pre-IPO analyst coverage in China. For example, cross-ownership is quite common among the investment banks and brokerage firms in China. One investor can have a control ownership of investment bank A while having a significant ownership in investment bank B (but the regulations prevent the same investor from having controlling ownership of two securities firm at the same time). We find that analysts affiliated with B are more likely to provide coverage for IPOs underwritten by A, and B’s analysts’ earnings forecasts for A’s IPOs are also more optimistic, everything else being equal.
Analysts’ reputational effects, however, seem to dominate the agency concerns for pre-IPO research coverage. The positive associations of analyst coverage and earnings forecast optimism with offer price revisions and initial returns remain statistically and economically significant after we remove the reports from analysts connected with an IPO due to cross-ownership. When we decompose analyst coverage and earnings forecast optimism into predicted and residual components, it is the unexpected, residual components of analyst research that drive the positive association between analyst research and IPO pricing. The residual components of analyst research are more likely to come from analyst information production instead of their hyping efforts. The latter result thus suggests that analysts gain their influence on IPO pricing through their information production efforts.
Furthermore, if analysts are hyping a stock before it goes public, with the offer price driven to a value above the stock’s fundamental value, we should observe a negative relation between the long-run returns on the IPO and the amount of analyst coverage and/or optimistic earnings forecasts. We do not find such a negative relation, whether we examine one-year, two-year, or three-year buy-and-hold market-adjusted returns after the IPO. We thus conclude that pre-IPO analyst coverage is in general providing information to investors, rather than merely hyping the stock.
In the U.S., research coverage is permitted before the IPO by unaffiliated analysts, just as is the case in China. Although this coverage is permitted in the U.S., it is virtually nonexistent. In contrast, in China coverage by unaffiliated analysts before the IPO is common. We conjecture that the difference is due to two reasons. First, in China, individual investors dominate the IPO market, unlike the U.S. Second, in the U.S. shares in an IPO are generally only available to investors who have an account at an underwriter, whereas in China investors apply for shares on-line and can have an account at any brokerage firm. When there is oversubscription, as is almost always the case, shares are allocated on a pro rata basis. Thus, investors with accounts at an unaffiliated brokerage firm have a demand for analyst coverage of a forthcoming IPO, and these brokerage firms supply the coverage.
The complete paper is available here.
