What to Know About an IPO

Initial public offerings (IPOs), particularly those involving familiar companies, often draw considerable attention from investors. In this article, we highlight some of the IPO investment hurdles faced by market participants and delve into the performance of IPO firms. We find that the Fama/French five-factor model largely explains the returns of hypothetical portfolios consisting of recent US IPOs. As a group, they have behaved like small growth, low profitability, high investment stocks. As such, they have underperformed the broad US market in their first year.

IPOs are commonly associated with outsized stock returns on the first day shares become available, although these returns may not be attainable by all investors due to the allocation process. Ritter (1987) shows initial trading prices typically exceed the IPO offering price, and the author documents an average first-day return of 17.9% from 1980 to 2018.¹ However, accessing these first-day returns requires an allocation from the underwriting banks. Studies have documented an adverse selection problem associated with IPOs: those with poor first-day returns have generally been easier to obtain in advance, while those with good first-day returns have usually been reserved for certain clients of the underwriting banks. For example, Reuter (2006) studied US offerings between 1996 and 1999 and found that business relationships with lead underwriters increased investor access to underpriced IPOs. More recently, Jenkinson et al. (2018) studied IPOs in the UK between 2010 and 2015 and found that most major investment banks, when making allocation decisions, favored clients that produce the most revenue.

Aftermarket support from the underwriting banks may also impact returns in the period immediately after the IPO. Ellis et al. (2000) examined 559 IPOs issued between 1996 and 1997 and showed that lead underwriter market activity represented more than 40% of initial trading volume in their sample. Hanley et al. (1993) studied IPOs from 1982 to 1987 and observed that prices tended to fall following the estimated conclusion of underwriter pricing support activities.

Given that many investors may have limited access to IPO allocations, this study focuses on the performance of IPOs in the secondary market. How do IPOs perform in their first year? One potential performance headwind for IPOs is the expiration of lockup agreements. Generally, a large percentage (typically 50% or more) of the IPO shares held by insiders are subject to lockup provisions that prevent such insiders from selling shares on the open market. When the lockup agreements expire, usually 6 to 12 months after the initial offering, these shares may be sold in the marketplace, creating a liquidation event that puts downward pressure on the stock price. Field and Hanka (2001), Brav and Gompers (2003), and Bradley et al. (2001) reported negative cumulative abnormal returns, defined as the stock’s return adjusted for the market return over the same period, over the three-day window surrounding the lockup expiration date for IPOs issued from 1988 to 1996.²

The sample in our study consists of 6,362 US IPOs that occurred between January 1991 and December 2018 for which data are available.³ Exhibit 1 shows the annual frequency and market cap distribution of IPOs among firm size groups. The period from 1991 to 2000 is characterized by a relatively high IPO frequency rate of 420 per year and is followed by a less active 18-year period during which the rate falls to 120 IPOs on average per year. Although the number of IPOs has declined, the average IPO offering size was almost three times larger over the most recent period, as compared to the initial 10 years in the sample.

Most IPOs fall into the small cap size group, defined as firms that fall below the largest 1,000 US-domiciled common stocks at the most recent month-end. Large cap and mid cap IPOs represent 24% and 19%, respectively, of total capital raised through IPOs.

We evaluate IPO returns by forming a hypothetical market cap-weighted portfolio consisting of IPOs issued over the preceding 12-month period, rebalanced monthly.⁴ By design, this methodology excludes the initial first-day returns to avoid the adverse selection problem in the IPO allocation process. Exhibit 2 compares the returns of the IPOs to the returns of the Russell 2000 and 3000 indices over the full sample period as well as two subperiods covering 1992 to 2000 and 2001 to 2018. We find IPOs underperformed the Russell 3000 Index in both the overall period and sub-sample periods. For example, IPOs generated an annualized compound return of 6.93%, 13.63%, and 3.74% in the overall, initial nine-year and final 18-year sample periods, respectively, as compared to 9.13%, 15.70%, and 5.98% for the Russell 3000 Index over the same time horizons. In comparison to the Russell 2000 Index, IPOs underperformed in the overall period (6.93% vs. 9.02%) and the 2001–2018 (3.74% vs. 7.29%) sub-period and outperformed (13.63% vs. 12.56%) from 1992 to 2000. IPO returns also exhibited higher levels of volatility relative to the Russell 3000 and 2000 indices.

Known drivers of returns largely explain the underperformance of IPOs. Exhibit 3 shows that the t-statistics for the intercept are below two for the full period and both sub-periods, suggesting IPO returns are well described by the Fama/French five-factor model. The results also indicate IPOs underperformed the market because, as a group, they have behaved like small growth, low profitability, high investment stocks, which tend to have lower expected returns than the market (Fama and French, 2015).

Investors considering IPOs should be aware of potential adverse selection and post-offering activities, such as underwriter pricing support and the expiration of insider lockup periods. In addition, we show IPOs generally underperformed the broader market benchmarks from 1992 to 2018, as well as in sub-periods with relatively high and low IPO activity.