Many companies that plan initial public offerings set
lock-up agreements--pacts that prevent a company's executives,
stockholders and option holders from selling shares during a specified
period without the express consent of the IPO underwriter.
The purpose of such a covenant--at least in theory--is to limit trading
volatility and maintain a loyal, stable shareholder base.
Essentially, underwriters who put up cash to buy shares in an offering are
looking to avoid a massive share dump once the deal is done.
Since 1996, more than 90 percent of IPOs have adopted some type
of lock-up agreement.
The trend is to include a lock-up agreement within a company's
prospectus, but there has been a decline in the average number of
days that the agreement can remain in effect. Since 1990, a typical lock-up
agreement totaled nearly 250 days, or about eight months. Insiders and executives that participated in an IPO in June 1997, for example, wouldn't be able to sell shares until February 1998.
In 1994, the typical lock-up agreement was 275 days, or about nine
months. After that, the average length of lock-up periods began to decline.
In 1999, the average lock-up agreement fell to less than 180 days.
This reduction in the number of no-trade days corresponded with an
unprecedented appreciation in IPOs on their opening trading days. Some
market observers maintain that it is to the advantage of insiders to shorten
the lock-up period to capitalize on the enormous price gains of many new
It should come as no surprise, then, that as IPO prices have risen
significantly, there has been a corresponding decline in the average duration
of lock-up agreements.
Executive compensation also has been a factor in shortening the length of
lock-up periods. Since many employees of Internet start-ups are often wooed
with stock options to join the firm, it's not uncommon to see an executive
from a successful newly public firm as a millionaire on paper but a
pauper at the bank. Shortening lock-up periods allows employees to readily cash
Conversely, there are arguments that lock-up agreements aren't designed to
keep insiders from selling early but are intended to limit the available supply of stock
on the open market, artificially driving up the share price and locking out
For example, the results of a study by Michael McBain and David Krause
published in a 1989 article in the Journal of Business Venturing maintained
that an IPO stock price is directly related to the proportion of insider
ownership. The premise is that lower insider ownership indicates higher
risk and adversely affects share price. Thus, as long as insiders
maintain their level of ownership, outsiders will have no choice but to bid
Research has shown
|IPOs with lock-up provisions|
| ||IPOs with lock-up provisions* ||Percentage of all IPOs ||Avg. days before expiration|
|1990 ||123 ||71.5 ||258|
|1991 ||328 ||89.4 ||254|
|1992 ||483 ||93.4 ||255|
|1993 ||627 ||88.7 ||254|
|1994 ||536 ||87.9 ||275|
|1995 ||518 ||89.3 ||248|
|1996 ||803 ||92.1 ||239|
|1997 ||579 ||91.8 ||234|
|1998 ||363 ||97.1 ||226|
|1999 ||461 ||92.2 ||184|
* excludes closed-end funds
Source: Securities Data Corp.
that IPOs with the shortest lock-up periods don't necessarily
outperform those issues with longer lock-up agreements. Several IPOs in
1999, including Bluestone Software, Salon.com and Fashionmall.com, came to
market with expiration dates of just 90 days or less.
In nearly every case, those IPOs with shorter lock-up agreements
didn't fare better than the typical IPO.
For example, Salon.com--which priced at $10 per share--had a 90-day lock-up
period in force. However, by the expiration day the shares had already
dropped more than 40 percent from their first day of trading.
IPOs with average lock-up agreements of 90 days gain approximately 13.1
percent by the time the lockup expires. That compares to a 26.2 percent
gain, on average, for IPOs whose lock-up agreement came to an end 180 days
after an offer date.
These results may be affected by the high number of IPOs with 180-day
lock-up periods, compared to those with shorter durations. Nonetheless,
several facts are irrefutable.
First, the average length of a lock-up agreement has been declining.
Second, the number of IPOs with 90-day or shorter lock-up periods has
venture-backed IPOs typically will have shorter lock-up periods than those
not backed by such funding.
For example, of the 77 IPOs that were supported by venture capital in 1998,
the corresponding lock-up periods averaged 184 days, compared to an average of 239 days
for IPOs without funding. The following year, VC-supported IPOs
averaged a lock-up period of 177 days, while others averaged 192 days.
One reason the length of lockups has dropped is that venture capital
firms and institutional owners are less willing to keep their full
investment intact for any extended period. As one money manger noted, the
current IPO market is one where venture funds must quickly return some of
their gains so partners can churn out more deals to remain competitive.
In part, the rapid appreciation of IPOs in the past few years has made it
imperative for institutions to "lock in" gains so they can document a strong
return to their own investors and complete the cycle--gather more capital to
fund more IPOs. Hence, the lock-up period gets shorter as
institutions look to bolt from their investments and seek the next deal.
But what happens after a lock-up period expires? An examination of recent
IPOs shows that among those priced in 1999 with lock-up agreements of 180
days, the average gain at the time of expiration was approximately 85
Since that time, however, those issues with expired lock-up provisions have
collectively registered an average gain of over 126 percent since their
offer date. This indicates that removing a lock-up agreement does little to
diminish investor support.