The frenzy and tumult associated with initial public offerings (IPOs) by young, high-tech companies during the Nasdaq bull market of 1998 to 2000 first fascinated America and then, as the market soured, provoked massive recriminations, including enforcement proceedings by the Securities and Exchange Commission (SEC) and even a criminal investigation by the United States Attorney for the Southern District of New York. Although the criminal investigation was ultimately dropped, the SEC has negotiated a $100 million settlement from Credit Suisse First Boston (CSFB) of charges brought by the SEC and the National Association of Securities Dealers (NASD) that the firm demanded excessive or unlawful brokerage commissions in return for IPO allocations. The SEC also continues to investigate allegations that various underwriters manipulated stock prices by tying IPO allocations to mandatory purchases of the same stock in the secondary market (a practice known as "laddering"), and numerous private class actions are pending that seek damages for investors who purchased IPO stocks in the secondary market.

But what do these investigations prove, beyond that some firms may have broken the law? For many, the characteristics of IPOs that most attracted public and media attention -- namely, the intense competition to receive allocations of shares in "hot" or oversubscribed offerings and the extraordinary first-day price spikes when shares sometimes rose 500 percent (or even more) above their offering price during the first day of trading -- also indicate that something is seriously wrong, or at least dysfunctional, within the IPO market.

Far less agreement exists, however, on what, if anything, is wrong. Some see the late 1990s era in moralistic terms as reminiscent of the period preceding the 1929 crash, when stock promoters foisted worthless securities on an unsuspecting public. Others view the same evidence as demonstrating that the IPO market is economically dysfunctional and favors the interests of financial intermediaries over those of young companies seeking to raise capital. Finally, defenders of the system maintain that the recent public outcry only repeats the familiar pattern of searching for a scapegoat whenever the stock market crashes. They argue: "If ain't broke, don't fix it."

· The First-Day Price Spike Puzzle

To resolve (or, at least, understand) these differing perspectives, it is useful to begin with the most salient fact about IPOs: namely, the first-day price spike. During 1999 and early 2000, the average IPO rose roughly 100 percent from its initial offering price to the close of trading on the first day. By definition, price spikes occur because the offering is oversubscribed -- that is, the underwriters have solicited "indications of interest" from potential buyers amounting to many times the number of shares that the issuer (i.e., the company "going public") wishes to sell. Often, the underwriters may solicit non-binding offers representing five or even ten times the shares that are to be sold. The underwriters "build their book" in this fashion in part because all offers to purchase by investors are legally nonbinding and revocable up to the point at which the SEC approves the issuer's registration statement and declares it "effective." Hence, the underwriters know that even a seemingly healthy 2:1 ratio of offers-to-buy to shares-to-be-sold could suddenly evaporate. Rationally, they wish to maximize this oversubscription ratio as a protection against a sudden shrinkage in demand if a substantial group of buyers were to back out at the last minute.

Understandable as this motivation is, it results in the underwriters possessing material nonpublic information, which they typically do not share with either the issuer or the public investors: namely, the size of this oversubscription ratio. For example, if this ratio were 7:1, the underwriters could reasonably expect that the offering would be a "hot" one with a sharp price spike in the immediate aftermarket, as investors who offered to buy the stock but did not receive an allocation (or were allocated less than they wanted) were forced to buy the stock after it started trading.

· The Allocation Issue

This knowledge faces the underwriters with a critical choice: They could prorate the stock in a "hot" or oversubscribed offering among all customers (or use some other equitable allocation formula), or they could allocate the stock largely to their preferred customers. Almost universally, underwriters choose the second option. Why? The simple answer is that large institutional investors implicitly pay for receiving priority in the allocation of "hot" offerings by directing their brokerage business to the major underwriters. For example, a large mutual fund may trade millions of shares in the secondary market each year. It could direct this brokerage business to a cheap discount broker, or it could negotiate a somewhat higher commission rate with a broker dealer that was also a major underwriter in return for a priority in the latter's IPO allocations. Some believe that there is a known "going rate" -- i.e., a minimum amount of brokerage business necessary to obtain an IPO allocation. But even if no such set rate exists, norms of reciprocity have long characterized the underwriting business. Indeed, using allocations to obtain brokerage business is probably what makes the IPO business very profitable for major underwriters, because for many years the major underwriters have all charged approximately the same underwriting commission (roughly 7 percent of the total offering -- a seemingly non-competitive price that has at times concerned the Justice Department's Antitrust Division, even if no conspiracy has ever been proven).

Retail investors may resent their relative exclusion from IPOs, but no rule of the SEC or the NASD requires that every customer receive an equal opportunity to buy "hot" stock. Indeed, across other fields of business, merchants characteristically reserve goods in scarce supply for preferred customers.

· Pushing the Envelope

The competition for IPO allocations intensified in the late 1990s, precisely as the first-day price spikes became larger and larger. As IPO allocations became more valuable, practices changed in several respects.

First, newer varieties of institutional investors -- most notably, hedge funds -- sought to obtain IPO allocations. Typically, they had paid far lower aggregate annual brokerage commissions to underwriters than had much larger mutual funds, but as the IPO market heated up in the late '90s the hedge funds were willing to compensate underwriters for this shortfall by paying considerably higher brokerage commissions on a per-share basis. For example, some hedge funds allegedly paid extraordinarily high commissions on individual transactions (say $50,000 on a $500,000 transaction) in order to earn a credit that could be applied to IPO allocations. To federal prosecutors who for a time were investigating these transactions, this looked suspiciously like commercial bribery: that is, exchanging a cash payment for an allocation.

Second, some underwriters allegedly began asking institutional clients to share their first-day trading profits with the underwriter (for example, by rebating one-third of their first-day profits). These practices did violate SEC or NASD rules, which preclude brokerage firms from sharing in their clients' profits or charging commissions in excess of 5 percent. The irony here is that these rules were designed to protect unsophisticated retail customers from overreaching by brokers, whereas the much more sophisticated hedge funds arguably did not need or want such paternalism (because if they could not pay such "excessive" commissions, they would likely be denied or reduced in their IPO allocations).

· Who Is the Victim?

Although allegations of bribery and suspiciously large brokerage commissions have dominated the media's coverage of the IPO allocation process, two other injuries may be more serious, if less visible. First, the corporate issuer may suffer the clearest injury when there is a substantial first-day price spike. Assume a first-time issuer goes public at a price of $10 per share (this means in reality that the underwriters buy the stock from the issuer at $9.30 and resell it to their preferred clients at $10). The first trade in the secondary market on the day of the offering occurs at $15 per share (a 50 percent runup), and by the end of that first day, the stock has risen to $45 per share (a 350 percent increase). If one million shares are now outstanding, the issuer's market capitalization at the end of the first day is thus $45 million -- but only $9,300,000 has been raised by the issuer. In other words, approximately 20 percent of the value of these securities has gone to the issuer to fund its growth and expansion, while 80 percent has gone to financial intermediaries. If the purpose of the equity market is to permit companies to raise capital to fund their economic expansion, this is a very costly form of financing.

Why doesn't the issuer object to this underpricing and insist that the underwriters price heavily oversubscribed offerings higher? For example, if the offering were oversubscribed by, say, a 7:1 ratio at a price of $10 per share (as in the preceding example), the issuer would realize that the underwriters could probably sell out the offering at $15 or even $20 per share, thus realizing more capital for the issuer and less profit for speculators. This puzzle of why so much money is seemingly "left on the table" by the issuer has long fascinated financial economists, who have found that the average IPO is underpriced by $9.1 million (an amount roughly twice the fees of the underwriters).1 The answer to this puzzle cannot be that IPO issuers are systematically stupid or naïve because the typical board of directors of a company going public usually includes several sophisticated venture capitalists.

Several different factors may explain why the underpricing of IPOs has long been a pervasive phenomenon. First, both issuers and underwriters face very high liability under the federal securities laws if the stock price in an IPO declines below the initial offering price. For this reason, they may consider it safer to underprice the offering to a degree in order to avoid the special standards of liability that apply to registered public offerings.

Second, both the issuer's management and the underwriter may face conflicts of interest that lead them to knowingly accept underpricing. Obviously, the more the offering is underpriced, the more valuable the IPO allocations become and the more the underwriters can demand that institutions use their brokerage services in return for them. For the issuer's management, the conflict is more complex and lies in the fact that they typically cannot sell their own shares in the company until six months after the IPO. Almost uniformly, underwriters negotiate a "lockup agreement" under which insiders (both management and venture capitalist shareholders) agree not to sell any of their shares until the expiration of a six-month lockup period that begins on the offering date. This both assures investors that management is not "bailing out" of the company and also intensifies the first-day price spike (because in the typical case the insiders own a majority of the stock and thus the lockup substantially restricts the short-term supply of stock available in the secondary market). Given that its own stock is locked up, management's personal focus may be on the likely value of the stock in six months time when they can sell their own shares. Both the conventional wisdom and the empirical research holds that a sharp first-day price spike is the best means of maximizing this later value because it creates a momentum in the stock's price that builds up to the lockup expiration date. That is, the greater the first-day price spike, the more that the firm will attract the attention of securities analysts and maximize its share value as of the end of the lockup period.2 Nonetheless, from a public perspective, the social price of benefitting management in this fashion by intentionally underpricing the stock is that the firm raises less capital and at a more expensive cost.

· Other Victims: Retail Investors

Another constituency may be even more adversely affected by current practices. This group consists of those investors who do buy in the secondary market, either on the first trading day or shortly thereafter. Typically, these are persons who did not receive an IPO allocation (or only a small one). The best known empirical regularity about IPOs is that the issuer's stock price tends to decline at some point during the year following the initial price spike, often to a level below the initial offering price. Thus, those who buy in the secondary market immediately following the IPO tend to lose.

Although this typical post-offering price decline may seem surprising, the reasons underlying it are logical. First, there is a systematic imbalance between supply and demand in the period following the IPO, with supply being deliberately restricted. Supply is constrained, first, by the lockup agreements that keep management's and the other insiders' stock off the market, and second, by "anti-flipping" rules. "Flipping" refers to the practice of quickly reselling IPO stock in the immediate aftermarket, often after a holding period lasting no more than a few days -- and sometimes much less. Underwriters dislike flipping because it creates downward price pressure on the stock (and disappoints the issuer's management, their key constituency). Thus, they restrict flipping by telling most investor clients that if they flip, they will be denied the right to purchase shares in future IPOs for at least a defined period (put into the "penalty box," in the vernacular). Brokers whose clients flip may also be subjected to financial penalties imposed by the underwriter by contract.

These anti-flipping rules are not generally applied to institutional investors, who simply have too much economic leverage to be disciplined in this fashion. As a result, in the immediate aftermarket, only institutions are freely permitted to sell, and hence the imbalance between supply and demand can become extreme, thereby contributing to the intense character of the price spike.

Although anti-flipping rules have been upheld by the courts (with the SEC's encouragement) and lockup agreements seem unquestionably valid and enforceable, other practices that developed in the late 1990s are legally suspect and probably unlawful. Chief among these are "tie-in" arrangements whereby the underwriter gives an institutional investor the allocation that it seeks (or most of it) on the condition that the institution will buy additional shares at an inflated price in the IPO aftermarket. For example, suppose in a hot offering in which the initial offering price will be $10 per share, the underwriter agrees to give an institutional investor the full 50,000 share allocation it wants if it agrees to immediately purchase 25,000 shares at a $15 price on the commencement of trading. Alternatively, the underwriter may demand that the institution buy 5,000 shares at each one-point price increment reached by the stock during the first day of trading. Such prearranged agreements to purchase stock at prices above the offering price artificially manipulate the market, in the SEC's view, and are the subject of a continuing SEC antifraud investigation. The victims of these practices are, of course, the retail investors who are buying at inflated prices that do not reflect the natural equilibrium of supply and demand.

· Prospects for Reform

Even if these fraudulent practices can be restricted, the IPO market will still remain one in which the short-term price is not the product of the normal interaction of supply and demand, because supply has been artificially restricted by lockup agreements and anti-flipping rules. As a result, price spikes in oversubscribed offerings should persist, and are likely to be followed by a price decline once these restrictions on supply lapse. For the present, investors have been sufficiently repelled by the bursting of the Nasdaq bubble that few IPOs are so oversubscribed as to produce any price spike. But in the future, the familiar pattern may reappear.

One market development could change this pattern. A few underwriters (most notably W.R. Hambrecht & Co.) are seeking to introduce an auction procedure for the pricing of IPOs. This procedure would price the stock at the highest price at which the entire offering could be sold. Thus, it would end underpricing and automatically give the retail investor an equal opportunity to participate in IPOs. Auction pricing faces strong opposition, however, both from underwriters (who want to exchange allocations for brokerage business) and from institutional investors (who obviously prefer to participate in underpriced offerings). Also, while auctions might ensure higher prices for issuers, there would be an increased risk of legal liability if the stock price dropped within the following year, and managements might receive lower prices for their own shares on the expiration of their lockup agreements. Thus, the prospect for sweeping changes in the market still seems remote.

What changes or reforms are likely from the SEC? Beyond enforcing existing laws, it is very unclear whether the SEC will seek to introduce any new rules or regulations. Inherently, the SEC's authority is to prevent fraud and require full disclosure. Thus, in all likelihood, it does not feel authorized to impose an auction mechanism on the IPO marketplace. Although the SEC probably also lacks authority to adopt an equal opportunity rule for the allocation of IPO stock, the NASD and the stock exchanges do have broader authority and arguably could impose such rules on underwriters. But any attempt to do so would likely set off a major lobbying battle.

What, if anything, should be unsettling to the average citizen in all these developments? Of course, indications of fraud or manipulation, particularly when committed within powerful and respectable institutions, should be disturbing. But the evidence is not yet in, and the extent of such misconduct cannot be estimated reliably at this point. The more serious policy problem surrounds the question of whether the contemporary IPO market fulfills the classic function of a market: namely, to link buyers and sellers at the lowest possible cost. Arguably, rather than linking capital-hungry corporations with potential providers of capital at low cost, the contemporary IPO market is designed more to benefit financial intermediaries (i.e., underwriters and those institutional investors who receive IPO allocations and then flip these shares in the short term). No consensus exists today on whether the IPO market is inefficient or otherwise imposes unnecessary costs on capital formation, but a lively and important debate seems likely to continue.

NOTES