IPO Listings: Where and Why?

Anne M. Anderson and Edward A. Dyl∗

According to most research, firms benefit from being listed on the New York Stock Exchange
(NYSE). Nevertheless, 224 of 640 firms that went public from 1993 through 2000 and were
eligible for a NYSE listing chose to list their stock on Nasdaq. We hypothesize that this choice may
be related to Securities and Exchange Commission (SEC) Rule 144. The rule regulates the sale
of restricted stock by limiting the amount of unregistered stock that can be sold by an individual.
We investigate the determinants of post-IPO sales of restricted stock, examine IPO firms’ listing
choices, and find evidence consistent with firms selecting Nasdaq to reduce the effect of the limits
on selling restricted stock imposed by the SEC’s Rule 144. Venture capitalists play an important
role in this listing decision.

Most research agrees that firms benefit from listing their stock on the New York Stock Exchange
(NYSE). Kadlec and McConnell (1994) and, more recently, Jain and Kim (2006) find that share
prices rise when firms move from Nasdaq to the NYSE. Nonetheless, not all firms that are eligible
to list on the NYSE choose to do so. Cowan, Carter, Dark, and Singh (1992) report that many
Nasdaq firms that qualify to list on the NYSE remain on Nasdaq. Corwin and Harris (2001) show
that from 1991 to 1996 more than 20% of the companies that qualified for listing on the NYSE
at the time of their initial public offering (IPO), decided to trade on Nasdaq instead.
We examine possible reasons for IPO firms to select Nasdaq over the NYSE, controlling for
characteristics of the firm such as its size, the volatility of its common stock, and whether or
not it is classified as a high-technology firm. We consider the motives of the founders, venture
capitalists, and other pre-IPO investors who choose where the stock will trade after the IPO. We
focus on whether some firms choose to list their stock on Nasdaq, rather than on the NYSE,
to obtain more advantageous regulatory treatment under Securities and Exchange Commission
(SEC) Rule 144, which governs post-IPO sales of restricted stock. Investors who own the stock
before the IPO can reduce the effect of Rule 144’s restrictions by choosing to have their stock
traded on Nasdaq rather than on the NYSE.
We investigate the relation between post-IPO sales of restricted stock and the listing decisions
of IPO firms from 1993 through 2000, controlling for the other factors that are relevant to the
exchange listing decision. We find that of the 640 IPO firms that are eligible to list on the
NYSE from 1993 to 2000, 35% choose instead to list on the Nasdaq, compared to only 23% of
the 438 IPO firms eligible for the NYSE that Corwin and Harris (2001) report for 1991-1996.
Firms that choose to list on the Nasdaq are smaller than those that opt for the NYSE and are
more concentrated in the high-tech sector. Of the firms that choose Nasdaq, 64% have sales of
restricted stock during the two-year period following the IPO, compared to only 39.8% of the
firms that choose the NYSE. Post-IPO sales of restricted stock are higher for volatile firms, for
venture capital (VC)-backed firms, and for firms go public after 1997, but not for firms that sell

We appreciate comments from Paul Bennett, Bruno Biais, Laura Field, Frank Hathaway, Jean Helwege, and Kathy Kahle,
and earlier input from Laurie Krigman. Data regarding sales of restricted stock are from a research project funded by the
New York Stock Exchange (Anderson, Dyl, and Krigman, 2005).

∗
Anne M. Anderson is an Assistant Professor of Finance at Lehigh University in Bethlehem, PA. Edward A. Dyl is the
Sheaf/Neill/Estes Professor of Finance at the University of Arizona in Tucson, AZ.
Financial Management • Spring 2008 • pages 23 - 43
24 Financial Management • Spring 2008
more secondary shares in the IPO, for firms that went public in hot markets, for firms that have
larger post-IPO stock price appreciation, or for firms with greater top management ownership.
When we use logistic regressions to model firms’ listing decisions and control for firm size,
underwriter quality, and membership in a technology industry, we find that venture capitalists are
a major influence behind firms’ choice of Nasdaq instead of the NYSE. We also use the number
of restricted shares sold during the two years following the IPO as a general proxy for pre-IPO
investors selling intentions, and find that post-IPO stock sales are strongly related to the listing
decision and are higher for firms that choose Nasdaq. Our interpretation of these results is that
venture capitalists and other pre-IPO investors who intend to sell more shares are motivated to
choose Nasdaq. This behavior facilitates the later sale of their restricted stock in the firm without
their having to undertake an SEO to comply with the provisions of the Securities Act of 1933.
The remainder of the paper proceeds as follows. In Section I, we provide background informa-
tion on SEC Rule 144. In Section II, we provide information about the data used in our analysis.
In Section III, we examine the determinants of restricted stock sales during the two years follow-
ing the firm’s IPO. Section IV reports our findings regarding the listing decision and post-IPO
restricted stock sales, and Section V concludes.

I. Background

The Securities Act of 1933 (the Act) requires that all securities being distributed to the public,
either by the issuer through an IPO or by an issuer and/or investors through a secondary distri-
bution, be registered with the SEC. Rule 144 exempts certain transactions from this registration
requirement. In particular, compliance with the rule provides a safe harbor that permits individ-
ual investors who own restricted stock to sell it to public investors. Absent an exemption, all
offers or sales of securities to public investors must comply with the registration and disclosure
requirements of Section 5 of the Act. 1 Rule 144 regulates how many shares of restricted stock
can be sold within specified periods after an IPO. The underlying rationale for the rule is to
enable individuals to conduct day-to-day trading transactions involving unregistered stock, which
are permitted under Section 4(1) of the Act, but to prohibit firms and/or investors from making
secondary distributions of large amounts of unregistered stock in violation of the Act.
An IPO usually includes only 20% to 30% of the total number of shares that the firm has
outstanding. Typical holders of the remaining 70% to 80% of the shares which are unregistered, and
are therefore restricted shares, are the firm’s early stage investors, including company founders,
other employees, venture capitalists, and mezzanine investors. Rule 144 specifies that during any
three-month window, the number of restricted shares that can be sold by an individual is limited
to the greater of:

1. One percent of the shares outstanding.

2. The average weekly trading volume in the security calculated over the four calendar weeks
preceding the week the notice of sale is filed on SEC Form 144.

We refer to these two provisions of Rule 144 as the One Percent Restriction and the Trading
Volume Restriction, respectively. Rule 144 also requires that restricted shares be held for a
minimum of one year from the time that they were originally acquired. Before 1997, this minimum
holding period was two years. We note that when shares have been owned for two years or more,
1
See Steinberg and Kempler (1988-1989).
Anderson & Dyl • IPO Listings: Where and Why? 25
no volume restrictions apply to nonaffiliated persons (essentially, noninsiders) who own restricted
shares. Insiders are always subject to volume restrictions.
The provisions of Rule 144 influence the speed with which pre-IPO shareholders can diversify
their personal portfolios by selling stock to public investors without registering the shares as
part of a secondary equity offering (SEO). Evidence suggests that this limitation is important
to shareholders. Kahl, Liu, and Longstaff (2003) find that the economic costs of such liquidity
restrictions to the firm’s entrepreneurs and other original shareholders can be sizeable. They show
that these liquidity costs can be of the same magnitude as those reported in the executive stock
options literature. Meulbroek (2001) finds that executive stock options are worth only 53% to
70% of their Black-Scholes value, and Hall and Murphy (2002) report that the value is 40%
to 60% of the Black-Scholes value. The post-IPO selling restrictions imposed by Rule 144 are
clearly important to pre-IPO shareholders of newly public firms who would like to sell a large
number of shares following an IPO, but who wish to do so without using an SEO.
In 1996, the NYSE petitioned the SEC to amend Rule 144 so that its trading volume standard
would operate comparably in dealer and auction markets. In a letter dated July 9, 1996, the NYSE
stated that “the double counting of trading volume in dealer markets results in a much higher
base figure for those markets from which to calculate permitted sales of restricted securities
pursuant to Rule 144. This enhanced ability to sell restricted securities listed on dealer markets
places auction markets at a competitive disadvantage in attracting new listings.” The SEC denied
this petition (in a letter dated February 19, 1997), stating, “The Petition provides no evidence to
support the alleged anticompetitive effects of dealer market trading volume on the ability of the
NYSE to attract new listings.”
If enabling pre-IPO investors to diversify their portfolios is an important objective when firms
go public, why don’t these investors simply sell more of their shares as a part of the IPO itself?
Investment banks dissuade pre-IPO investors from selling a material portion of their holdings in
the IPO because insider selling in an IPO conveys negative information about the value of the firm
(Leland and Pyle, 1977). A recent study by Ang and Brau (2003) underscores the concern about
insiders selling stock in IPOs. They find that insiders devise elaborate ways to conceal the extent
of their selling in IPOs. Insiders use the widely circulated original prospectus to underreport the
number of shares that they plan to sell and later use an obscure amendment to communicate
the true number of shares they will sell. Such legerdemain notwithstanding, investment banks
frequently dissuade insiders from selling any secondary shares in the IPO to avoid a negative
signal, so the act of going public does not per se solve insiders’ exit problems.
Lockup agreements, signed with investment bankers at the time of the IPO, also restrict pre-
IPO investors from selling shares immediately after an IPO. Such lockup agreements prevent
pre-IPO owners from selling shares for a specified period—typically six months—following
the IPO, and they generally apply to most of the shares not sold in the IPO. Field and Hanka
(2001) examine 1,948 lockup agreements from 1988 through 1997 and find that trading volume
increases permanently when lockup agreements expire, with a significant negative abnormal
return. Bradley, Jordan, Yi, and Roten (2001) find a similar result for 2,529 lockup expirations
from 1988 to 1997, and they also report that the losses are concentrated in firms with VC backing.
Like lockup agreements, Rule 144 provides public investors with some assurance against
pre-IPO investors’ selling a large amount of stock following the IPO. Goergen, Renneboog, and
Khurshed (2006) liken Rule 144 in the US to the mandatory lockup agreements found in European
IPO markets. The listing decision enables pre-IPO investors to signal a greater willingness to be
bound by the limitations of Rule 144. When they announce that the firm will trade on the NYSE,
they convey information about their selling intentions that could reduce the agency costs inherent
in external equity financing.
26 Financial Management • Spring 2008
Conversely, pre-IPO investors can reduce the effect of Rule 144’s restrictions by choosing to
have their stock traded on Nasdaq rather than on the NYSE, albeit at the cost of sending a possibly
negative signal regarding their selling intentions. Firms can determine how trading volume will
be measured for regulatory purposes, depending on whether they choose to list their shares on an
auction market, such as the NYSE or American Stock Exchange (Amex), or on a dealer market
such as Nasdaq. The extent to which Rule 144 limits post-IPO sales of restricted shares by insiders
is thus partially at the discretion of the managers of firms that have a choice between Nasdaq and
the NYSE.

II. Data
Using the Securities Data Corporation (SDC) New Issues database, we identify firms that go
public from 1993 through 2000. Excluding closed-end funds, REITS, ADRS, rights offerings,
privatizations, spin-offs, and issues with an offer price less than $5.00, there are 3,889 IPOs
during this period, 85% of which trade on Nasdaq following their IPOs.

A. Sample Selection
Because we focus on the listing choices of firms when they go public, we identify the IPO firms
for our sample period of 1993-2000 that actually have a choice between trading on Nasdaq and on
the NYSE. We screen the sample period’s population of IPOs for firms that were eligible to list on
the NYSE at the time of the IPO. The NYSE’s listing requirements, which changed periodically
during our sample period, primarily concern the firm’s size, the number of public shares, and the
firm’s earnings history. (See the Appendix for a summary of the NYSE listing guidelines from
1993 to 2000.) We otain the balance sheet and income statement data used to determine whether
or not the firm was eligible for listing on the NYSE primarily from Compustat; when the data are
not available from Compustat, we obtain them from SEC filings. We use the NYSE and Nasdaq
to obtain the information to calculate the initial and annual listing fees that would have been
incurred by each firm.
The Vickers Stock Research Corporation provides us with the information about sales of
restricted shares during the two years following each firm’s IPO. Data on stock prices and trading
volumes during the two years following the IPO are from the Center for Research in Security
Prices (CRSP).
Our final sample consists of 640 firms that qualify to be listed on the NYSE at the time of
their public offering. These 640 firms comprise only 16.5% of the 3,889 firms that went public
from 1993 through 2000, indicating that most IPO firms do not have a choice of trading venue.
Of this sample of 640, 416 (65%) elect to list on the NYSE and 224 (35%) listed on Nasdaq.

B. Characteristics of the Firms in Our Sample

Table I contains information about the 640 firms that qualify for listing on the NYSE at the time
of their IPO, classified according to the firm’s exchange listing following the IPO. We find that
firms that select the NYSE are much larger than those that list on Nasdaq, with a mean (median)
market capitalization of $1,481 ($494) million, compared to a mean (median) capitalization of
only $266 ($181) million for the Nasdaq firms. The mean and median offer prices for the NYSE
firms are $18.07 and $17.00, respectively, compared to mean and median offer prices of $14.58
and $14.00, respectively, for the Nasdaq firms. On average, the NYSE firms have significantly
more total shares outstanding than do Nasdaq firms, and they issue shares at higher prices. The
Anderson & Dyl • IPO Listings: Where and Why? 27
Table I. IPO Firms Classified According to the Listing Decision

This table shows information about 640 firms that go public between 1993 and 2000, and that meet the
NYSE’s listing requirements. The sample includes 224 Nasdaq firms and 416 NYSE firms. The last two
columns contain p-values from a t-test of means and a Kruskal-Wallis test of medians across exchange
listing.

Nasdaq Firms NYSE Firms Test of Test of

Means Medians
Mean Median Mean Median (p-Value) (p-Value)
Firm size (millions) $266 $181 $1, 481 $494 (0.0001) (0.0001)
Offer price $14.58 $14.00 $18.07 $17.00 (0.0001) (0.0001)
Shares outstanding (millions) 17.69 12.19 65.59 28.00 (0.0001) (0.0001)
IPO proceeds (millions) $57.61 $49.05 $208.76 $107.45 (0.0001) (0.0001)
Shares offered (% of shares 29.37% 27.83% 27.42% 23.11% (0.0002) (0.0001)
outstanding)
Secondary shares (% of 17.60% 13.38% 20.80% 0.00% (0.0268) (0.0001)
shares offered)
Listing fee difference (NYSE $0.092 $0.072 $0.237 $0.139 (0.0001) (0.0001)
minus Nasdaq) (millions)
Underwriter quality 7.98 8.10 8.77 9.10 (0.0001) (0.0001)
Stock price volatility (σ ) 3.41 2.15 2.26 1.88 (0.0036) (0.0815)
Post-IPO price increase 43% 32% 19% 15% (0.0001) (0.0005)
Age at time of IPO (years)a 15.47 11.00 24.92 11.00 (0.2762) (0.9154)
a
The founding year is available for only 219 of 224 Nasdaq firms and for 361 of 416 NYSE firms.

average proceeds of the IPOs are also larger for the NYSE firms, with mean (median) IPO proceeds
of $209 ($107) million for the NYSE firms compared to $58 ($49) million for Nasdaq firms.
In each group of firms, the number of shares offered in the IPO is a small proportion of the
total shares outstanding. The mean (median) proportion of shares offered in the IPO is 27.4%
(23.1%) of shares outstanding for the NYSE firms and 29.4% (27.8%) for Nasdaq firms. These
results are roughly consistent with those of Field and Hanka (2001), who report that for 1,948
IPOs that take place between 1988 and 1997, the average IPO sells 33% of the total number
of shares outstanding. Secondary shares average 20.8% of the shares offered in the IPOs of the
NYSE firms, compared to 17.6% for the Nasdaq firms, but the corresponding medians are zero
and 13.38%, respectively.
Since listing fees are a function of the number of shares the firm has outstanding, the number
of shares outstanding may be a factor in firms choosing to list on a given market. However, fees
are also higher on the NYSE than on Nasdaq for a firm of a given size. Listing fees vary from
year to year during 1993-2000. We calculate the listing fees that each firm would have incurred
in each market at the time of their IPO, and report the difference in fees between listing on the
NYSE and listing on the Nasdaq in Table I. The total fee we use for each market includes both the
initial listing/entry fee and annual fee for the first year. The table indicates that firms that choose
to list on Nasdaq incur a mean (median) fee that is $92,000 ($72,000) lower than if they had
listed on the NYSE, but firms that list on the NYSE pay mean (median) fees that are $237,000
($139,000) higher than if they had chosen to trade on Nasdaq. These amounts are small relative
to the size of the IPOs.
We report a measure of the “quality” of the lead underwriter for each offering, using the
Carter-Manaster (1990) rankings of underwriter prestige as updated by Loughran and Ritter
28 Financial Management • Spring 2008
Table II. Characteristics of the Sample of IPO Firms

This table shows the number and proportion of IPO firms that meet NYSE listing requirements and have the
characteristic indicated are shown for the sample firms and for subsets of the firms classified according to
listing choice. The test of means tests whether the proportion of Nasdaq firms with a particular characteristic
is different from the proportion of NYSE firms with that characteristic.
Firm Characteristic All Sample Firms Nasdaq Firms Test of NYSE Firms
Means
Number Percent Number Percent Number Percent (p-Value)
IPO firms 640 100% 224 35% 416 65% N/A
Venture capital backing 119 19% 74 62% 45 38% (0.0001)
High-tech firms 153 24% 87 57% 66 43% (0.0001)
IPO in a hot market 410 64% 144 35% 265 65% (0.8835)

(2004; see also Carter, Dark, and Singh, 1998). We obtain these data from Jay Ritter’s Web
site (http://bear.cba.ufl.edu/ritter/ipodata.htm). Table I shows that by this measure of underwriter
quality, the investment banks that underwrite the NYSE firms are of higher quality than are those
that underwrite Nasdaq firms.
The next two items in Table I are the volatility of the firms’ stock price, which we measure
as the standard deviation of the daily returns on the firm’s stock during the 100 trading days
following the IPO, and the post-IPO increase in the price of the firm’s stock, which we measure
as the percentage change in the stock price during the six months following the IPO. The Nasdaq
firms are more volatile than the NYSE firms, to a mean (median) standard deviation of returns
of 3.41 (2.15) compared to a mean (median) of 2.26 (1.88) for the NYSE firms. The Nasdaq
firms also show greater price appreciation during the six months following the IPO. The mean
(median) percentage price increase for the Nasdaq firms is 43% (32%) compared to a mean
(median) increase of 19% (15%) for the NYSE firms. The differences in the mean and median
volatilities of the firms’ stock are significant at the 1% and the 10% levels, respectively, and the
differences in the mean and median stock price increases are significant at the 1% level.
The final item in the table is the age of the firm at the time of the IPO. The mean age of the
NYSE firms is much higher than that of the Nasdaq firms (24.9 years compared to 15.5 years),
but the median ages of the NYSE and Nasdaq firms are the same (11 years), and neither the
means or the medians are significantly different.
Table II provides information about other characteristics of the IPO firms in our sample. As
noted earlier, 35% of the firms choose to trade on Nasdaq, and 65% choose to list on the NYSE.
Venture capitalists back only 119, or 19%, of the firms that qualify to list on the NYSE. However,
62% of these VC-backed firms choose to list on Nasdaq, but only 38% choose the NYSE,
proportions that contrast markedly with the overall sample. The proportion of VC-backed firms
that choose Nasdaq is significantly different from the proportion that choose the NYSE (p =
0.0001). There are 153 IPOs in high-tech industries, or 24% of the total sample, and the majority
of these firms (57% compared to 43%) choose Nasdaq. 2 This difference is also significant (p =
0.0001). Although firms with VC backing and high-tech firms both tend to choose Nasdaq, the
two sets of firms are not identical. Fifty of the high-tech firms (33%) are backed by venture
capitalists, compared to 17% of the full sample, and 43% of the VC-backed firms are high-tech
firms, compared to 24% of the full sample.

2
A breakdown of industry affiliation by exchange listing is available from the authors on request.
Anderson & Dyl • IPO Listings: Where and Why? 29
Of the IPOs in our sample, 410 (64%) occur in a so-called hot IPO market, which Helwege and
Liang (2004) define as a three-month period during which there are more than 30 IPOs per month.
The proportion of these firms choosing Nasdaq compared to the NYSE is indistinguishable from
the proportions for the overall sample of IPOs that qualify for an NYSE listing.

III. Determinants of Post-IPO Sales of Restricted Stock

Rule 144 limits an investor’s open market sales of restricted stock in any three-month period to
the greater of either 1% of the shares outstanding or the average weekly trading volume during
the four weeks preceding the sale. Whether the One Percent Restriction or the Trading Volume
Restriction is the binding restriction at a particular time depends on the sales of the firm’s shares
during the preceding four weeks. When a stock’s annual turnover rate during a particular four-
week period exceeds 52%, the average weekly trading volume in the stock will exceed 1% of the
shares outstanding. The average (median) annual turnover for the firms in our sample during the
two years following the IPO is 101% (79%) for the NYSE firms and 168% (144%) for the Nasdaq
firms. For each firm, we measure the average annual turnover during the two years following the
IPO by dividing the trading volume reported for each year by the number of shares outstanding
at the end of the year and then averaging the two values. However, trading volume varies greatly
from week to week, so annual turnover is not a good indicator of how often the sales of restricted
shares permitted by the Trading Volume Restriction exceed the sales allowed under the One
Percent Restriction. Annual turnover is an average, and a firm’s average weekly trading volume
can be extremely high during a few months and very low the rest of the time.
To provide a more precise estimate of the effect of the Rule 144 restrictions on the firms in
our sample, we calculate which restriction permits the greatest sales of restricted shares for 26
consecutive, nonoverlapping four-week trading periods during the two years following the IPO.
Table III shows the proportion of the time that each of the two Rule 144 provisions limits the selling
of restricted shares in our sample. The One Percent Restriction is the relevant restriction for the
average (median) Nasdaq firm 19% (12%) of the time, and the Trading Volume Restriction is the
relevant restriction 81% (78%) of the time. In contrast, the One Percent Restriction applies for the
average (median) NYSE firm 41% (38%) of the time and the Trading Volume Restriction applied
59% (62%) of the time. We attribute the greater importance of the Trading Volume Restriction
on Nasdaq to higher reported trading volumes, relative to the number of shares outstanding. The
higher reported trading volume is an artifact of the manner in which trading volume is reported
in a dealer market.
Figure 1 shows separately for Nasdaq and the NYSE firms the monthly trading volumes for the
firms in our sample during the two years following the IPO. Trading volume is very high for both
groups during the first month of trading, but by the third month of trading, the monthly volume
has declined by roughly two-thirds. The monthly volume is higher for the NYSE firms during the
first six months of trading, but Nasdaq firms have higher average volume for most of the rest of the
period. Most lockup agreements expire after six months, and there is a spike in Nasdaq volumes at
about this time. We suggest that the spike in volume and the subsequent higher monthly volumes
for Nasdaq firms reflect the sale of restricted shares by venture capitalists, founders, and other
insiders. There is no noticeable change in trading volume for the NYSE firms coinciding with
lockup expirations. Although the monthly trading volumes are relatively similar for the Nasdaq
and NYSE firms, turnover is much higher for the Nasdaq firms, reflecting the large difference in
shares outstanding between the two sets of firms.
30 Financial Management • Spring 2008
Table III. Percentage of Time a Restriction is Binding Under Rule 144

This table shows the proportion of time that each provision of Rule 144 is the binding constraint on sales of
restricted stock (i.e., permits the greatest sales of restricted shares) during 26 consecutive nonoverlapping
four-week trading periods in the two years following 640 IPOs from 1993 to 2000 where the firms qualify
for an NYSE listing at the time of the IPO.

1% Restriction Trading Volume Restriction

Mean Median Mean Median
Nasdaq firms 19% 12% 81% 78%
NYSE firms 41% 38% 59% 62%
p-value <0.001 <0.001 <0.001 <0.001

Figure 1. Post-IPO Monthly Trading Volume for Nasdaq and NYSE Firms

Figure 1 shows the monthly trading volumes for the firms in our sample during the two years following the
IPO.

A. Post-IPO Sales of Restricted Stock

We obtain our data on actual post-IPO open market sales of restricted shares for the firms in
our sample from SEC Form 144 filings from the Vickers Stock Research Corporation. Panel A
in Table IV summarizes these data. Of the 640 firms in our sample, 310 have sales of restricted
stock in the two years following their IPOs. The likelihood that a firm will have sales of restricted
stock during this period is higher for the Nasdaq firms than for the NYSE firms. More than 64%
Anderson & Dyl • IPO Listings: Where and Why? 31
Table IV. Sales of Restricted Stock during the Two Years Following the IPO

This table provides information about sales of restricted shares during the two years following each firm’s
IPO from filings of the SEC’s Form 144 obtained from the Vickers Stock Research Corporation. Data in
Panel A are for the 640 firms with IPOs from 1993 to 2000 that qualify for an NYSE listing at the time of
the IPO. Data in Panel B are for the 310 firms with IPOs from 1993 to 2000 that report sales of restricted
shares in the two years following the IPO. The last two columns contain p-values from a t-test of means and
a Kruskal-Wallis test of medians across exchange listing.

Nasdaq Firms NYSE Firms Test of Test of

Means Medians
(p-Value) (p-Value)
Panel A. N = 640 Firms

Number of firms 224 416 N/A N/A

Number of firms with restricted 144 166 N/A N/A
stock sales
Proportion of firms with restricted 64.29% 39.90% (0.0001) N/A
stock sales
Panel B. N = 310 Firms

Mean Median Mean Median

Sellers per firm 14.61 6 8.86 4 (0.0136) (0.0394)
Days with sales per firm 27.99 12.0 15.46 6 (0.0004) (0.0009)
Average shares sold per firm (million) 1.21 0.29 0.61 0.18 (0.0210) (0.1114)
Shares sold (% of IPO shares) 27.48% 8.86% 7.46% 2.80% (0.0001) (0.0001)
Share sold (% of restricted shares) 6.96% 3.03% 3.64% 1.04% (0.0001) (0.0001)

of the firms that choose Nasdaq have investors who sell restricted shares during the two years
following the IPO compared to only 40% of the NYSE firms. This difference is significant at the
1% level.
Panel B in Table IV shows that among the 310 firms that have restricted stock sales during the
two years following the IPO, Nasdaq firms have both significantly more sellers per firm and more
days with restricted stock sales than do the NYSE firms. The mean (median) number of sellers per
firm for the Nasdaq firms is 14.61 (6) sellers, compared to 8.86 (4) sellers for the NYSE firms,
and the mean (median) Nasdaq firm has 27.99 (12) days with restricted stock sales, compared to
15.46 (6) days for the NYSE firms. Investors in Nasdaq firms sell an average (median) of 1.21
million (290,000) restricted shares during the two years following the IPO, compared to average
(median) sales of 610,000 (180,000) per NYSE firm, even though the NYSE firms have a greater
number of restricted shares outstanding. The average (median) Nasdaq firm’s shareholders sell
restricted shares equal to 27.48% (8.86%) of the shares offered in the IPO, compared to an average
(median) of only 7.46% (2.80%) for the NYSE firms’ shareholders. As a percentage of restricted
shares outstanding, the average (median) proportions of restricted share sales are 6.96% (3.03%)
for Nasdaq firms and 3.64% (1.04%) for the NYSE firms. Overall, the pre-IPO shareholders of
firms that list on Nasdaq are much more active in selling their holdings following the IPO.

B. Tobit Model of Restricted Stock Sales

We examine the cross-sectional determinants of sales of restricted stock in IPO firms during
the two years following the IPO. We perform this part of our analysis irrespective of the listing
32 Financial Management • Spring 2008
choice. The dependent variable in our analysis of post-IPO open market sales of restricted stock
is the number of restricted shares sold during the two years following the IPO. The independent
variables in the basic Tobit analyses are the volatility of the stock, whether or not venture capitalists
are pre-IPO backers of the firm, the number of secondary shares sold in the IPO, the number of
shares outstanding, whether or not the IPO takes place before 1998, and whether or not the IPO
takes place during a hot IPO market. The rationale for using these independent variables is as
follows:

Volatility. Presumably, a major reason that an IPO firm’s founders and managers, venture
capitalists, and other pre-IPO investors might be more inclined to sell restricted shares following
an IPO is to diversify their portfolios. The riskier the firm’s stock, the greater will be this incentive
to diversify. Therefore, we include the volatility of the firm’s stock, which we measure as the
standard deviation of the firm’s daily stock returns during the 100 trading days following the IPO,
as an independent variable in the analysis. We hypothesize that post-IPO sales of restricted stock
are positively related to the volatility.
VC Backing. Black and Gilson (1998) suggest that IPOs provide an exit strategy for venture
capitalists and other private equity investors. Helwege and Packer (2005) find empirical evidence
that supports this proposition, and they conclude that a primary motive for firms going public
is that venture capitalists wish to liquidate their holdings. We include VC backing as a possible
determinant of post-IPO selling, and hypothesize that this variable is positively related to sales of
restricted stock during the two years following the IPO. We measure VC backing as an indicator
variable, equal to one if venture capitalists are pre-IPO investors in the firm and zero otherwise.
Secondary Shares. Primary shares are new shares issued as part of the IPO with the proceeds
going to the firm. Secondary shares are outstanding shares that are owned by pre-IPO investors
and are sold to the public as a part of the IPO. We hypothesize that the proportion of the IPO that
consists of secondary shares, rather than primary shares, will be negatively related to post-IPO
sales of restricted stock, because the investors who sell shares in the IPO have already diversified
their portfolios to some extent and, therefore, have less motivation to sell additional (restricted)
shares following the IPO. We measure this variable as the number of secondary shares sold in the
IPO as a percentage of total shares sold in the IPO.
Shares Outstanding. We use the logarithm of the number of shares outstanding as a scalar
variable because we measure the dependent variable in terms of the absolute number of shares
sold. We hypothesize that the number of restricted shares sold following the IPO will be positively
related to the total number of shares outstanding.
Pre-1998 IPO. In 1997, the minimum holding period required before restricted shares could
be sold changed from two years to one year. We include this variable in the model because the
change in Rule 144 in 1997 could have had an impact on either the decision to sell restricted stock
following the IPO and/or how many restricted shares are available for sale. Ceteris paribus, more
restricted shares will be available for sale after 1998 than before 1998, so we hypothesize that
post-IPO sales of restricted shares will be negatively related to whether or not the IPO occurred
after 1997. This variable is an indicator variable, equal to one if the IPO took place before January
1998, and zero otherwise.
Hot IPO Markets. In Table II, we show that 64% of the IPOs in our sample occur during a
hot IPO market. Lowry and Schwert (2002) examine cycles in the volume of IPOs and find that
successful IPOs result in high future IPO volume. If IPO firms that go public during hot IPO
markets are behaving opportunistically to take advantage of temporary overvaluation due to a
high demand for IPO stocks, then investors in these firms may be more inclined to sell their
shares after the IPO. Therefore, we hypothesize that post-IPO sales of restricted shares will be
Anderson & Dyl • IPO Listings: Where and Why? 33
higher for IPOs in hot markets. Following Helwege and Liang (2004), we define hot IPO markets
as periods with three consecutive months that have a moving average IPO count of more than 30
IPOs per month.
We also estimate regressions using two other measures of post-IPO sales of restricted stock as
dependent variables: 1) the presence/absence of any restricted stock sales during the two years
following the IPO (yes = one, no = 0), and 2) the number of individual sellers of restricted shares
during the two years following the IPO. The results are qualitatively the same as the estimates
reported in Table V.
Table V reports the parameter estimates for our regression analysis of restricted stock sales,
where the dependent variable is the number of restricted shares sold in the two years following
the IPO. We use a Tobit regression to estimate the parameters of the model instead of ordinary
least squares (OLS) because the value of the dependent variable in each regression equals zero
for 331 of the 640 observations in our sample. Using a Tobit regression is preferable to OLS
when the distribution of the dependent variable is censored in this manner.
We report the results of the basic regression using the six independent variables described
earlier as Model 1 in the table. Sales of restricted stock during the two years following the
IPO are positively related to the volatility of the firm’s stock and to the presence of venture
capitalists as pre-IPO investors, and the coefficients are significant at the 1% level. The number of
restricted shares sold is also directly and significantly related to the number of shares outstanding.
Conversely, restricted stock sales are lower for firms that go public before 1998, when the
minimum holding period for restricted stock was two years instead of the current one year. This
coefficient is significant at the 1% level. The coefficients on the proportion of secondary shares
offered in the IPO, whether or not the IPO occurred in a hot market, and the post-IPO stock
price increase, have the expected signs but they are not significant. In summary, excluding shares
outstanding, three of the independent variables are significant, each has the predicted sign, and
the pseudo R squared of the Tobit regressions is 13.80%. An OLS test using the same variables
yields parameter estimates that are essentially identical in magnitude and significance to the Tobit
results, with an R squared of 17.8%

C. Extensions of the Basic Tobit Model of Restricted Stock Sales

Here, we investigate whether either the post-IPO change in the stock price, the proportion of
the stock owned by top management, or the firm’s listing choice are related to open market sales
of restricted stock during the two years following the IPO.
Post-IPO Price Increase. It is plausible that the post-IPO performance of an IPO firm’s stock
price might influence post-IPO sales of restricted stock. As we see from Table II, on average,
the Nasdaq firms in our sample appreciate by 43% during the six months following the IPO and
the NYSE firms’ stock prices increase by 19%. The median price increases are 32% and 15%,
respectively. If the post-IPO appreciation in a firm’s stock price greatly exceeds the expectations
of the firm’s pre-IPO owners, this event might induce them to sell more shares than originally
intended following the IPO. We examine this possibility by adding Post-IPO Price Increase as
an independent variable in the model. We measure this variable as the percentage change in the
stock price during the six months following the IPO, because most restricted stock is subject to
lockup agreements that expire after six months. We report the results as Model 2 in Table V.
The coefficient on Post-IPO Price Increase has a positive sign, but it is not significant. Sales of
restricted shares are not related to post-IPO stock price run-ups.
Top Management Stock Ownership. Going public is only one event in a firm’s growth and
development. In addition to giving the firm access to the public market, it also provides early,
34 Financial Management • Spring 2008
Table V. Determinants of Restricted Stock Sales

The dependent variable is the number of restricted shares sold during the two years following the IPO. The
independent variables in the tobit regressions are: volatility, measured as the standard deviation of stock
returns during the 100 trading days following the IPO; secondary shares as a percentage of IPO shares; the
log of shares outstanding; the post-IPO stock price increase measured as the price at the end of six months
after the IPO divided by the offer price; the percentage of shares owned by the CEO and/or Chairman prior
to the IPO; and indicator variables denoting firms with venture capital backing, IPOs occurring before 1998,
IPOs during hot IPO markets, and IPOs that list on the NYSE.

Independent Variables Model 1 Model 2 Model 3a Model 4

∗∗ ∗∗
Intercept −0.5334 −0.5650 −1.9113∗∗∗ −0.5076∗∗
Volatility (σ ) 0.2474∗∗∗ 0.2452∗∗∗ 0.2549∗∗∗ 0.2408∗∗∗
Venture capital backing 0.5417∗∗∗ 0.5410∗∗∗ 0.8331∗∗ 0.4637∗∗∗
(yes = 1, no = 0)
Secondary shares (% of IPO) −0.1758 −0.1745 −0.1874 −0.1514
Shares outstanding (Log) 0.1657∗∗∗ 0.1669∗∗∗ 0.4892∗∗∗ 0.2414∗∗∗
Pre-1998 IPO (yes = 1, no = 0) −0.5521∗∗∗ −0.5611∗∗∗ −0.7261∗∗ −0.5363∗∗∗
Hot market (yes = 1, no = 0) 0.2240 0.2259 0.6330∗∗ 0.2327
Post-IPO price increase – 0.0309 – –
Top management ownership (%) – – 0.5061 –
NYSE listing (yes = 1, no = 0) – – – −0.3926∗∗∗
Pseudo R squared 13.80% 13.80% 13.52% 14.16%
∗∗∗
Significant at the 0.01 level.
∗∗
Significant at the 0.05 level.
a
Only 278 observations have information available about pre-IPO ownership.

temporary investors such as venture capitalists and mezzanine investors with the opportunity
to liquidate their holdings and move on to other investments. The positive average reaction to
the IPOs of the firms in our sample suggests that investors do not anticipate that the firm’s top
managers plan to liquidate their holdings, but an IPO undeniably provides controlling shareholders
with an opportunity to sell their restricted shares after the lockup agreement expires.
We investigate whether controlling shareholders are post-IPO sellers of restricted stock by
including Top Management Ownership as a variable in the model. We measure the variable as the
percentage ownership of the firm’s chief executive officer and chairman before the IPO. Model 3
in Table V shows the results of this analysis. We estimate Model 3 by using only 264 of the firms
in our sample, because data on top manager ownership are only available electronically for this
subset of the firms in our sample.
Stock Market Listing. Compared to the NYSE, the way Nasdaq reports trading volume means
that under the Trading Volume Restriction of Rule 144, stockholders of NYSE firms can sell
fewer restricted shares than can their counterparts in Nasdaq firms. Because of this difference,
sales of restricted shares may be lower for NYSE firms than for Nasdaq firms. We investigate
this possibility by adding a variable for the listing decision to the Tobit regression in Table V. The
NYSE Listing variable is an indicator variable equal to one if the firm lists on the NYSE, and
zero otherwise. We hypothesize that post-IPO sales of restricted shares will be negatively related
to an NYSE Listing.
We report the results of the regressions that include the firm’s listing choice in Model 4 in
Table V. The coefficient on the NYSE Listing variable has the predicted negative sign and is
significant at the 1% level. This relation is consistent with the proposition that the Trading
Volume Restriction of Rule 144 is more stringent for firms that list on the NYSE. Adding the
Anderson & Dyl • IPO Listings: Where and Why? 35
listing variable to the regression increases the pseudo R squared of the Tobit model from 13.80%
to 14.16%. The issue of causality is clearly a two-way street, but, ex post, listing on the NYSE is
associated with lower sales of restricted stock.
The relation between post-IPO sales of restricted shares is consistent with there being a greater
incentive to diversify one’s portfolio when the shares are riskier. The importance of pre- versus
post-1998 IPOs suggests that the change in the minimum required holding period for restricted
shares from two years to one year meant that more restricted shares were eligible for sale following
the IPO. Both of these results are as expected.
The results regarding venture capitalists and open market sales of restricted stock indicate
that venture capitalists are sellers of restricted stock following the IPO. This result is consis-
tent with findings by Field and Hanka (2001) and Bradley et al. (2001), who report significant
negative abnormal returns following the expiration of IPO lockup agreements, especially for
firms backed by venture capitalists. When lockup agreements expire, venture capitalists ap-
pear to be more aggressive sellers of shares than other pre-IPO investors. We do not have
specific information about the expiration dates of the lockup agreements in our sample, but
the overwhelming majority of lockup agreements expire six months after the IPO. In this
regard, recall that in Figure 1, we saw a spike in trading volume for Nasdaq firms around
month six.
In Figures 2 and 3, we examine monthly trading volume on each exchange separately for firms
backed by venture capitalists and other, non-VC-backed, firms. Figure 2 shows monthly trading

Figure 2. Post-IPO Monthly Trading Volume for NYSE Firms

Figure 2 compares monthly trading volume for firms with VC backing against those without VC backing
for NYSE firms for the two years following the IPO.
36 Financial Management • Spring 2008
Figure 3. Post-IPO Monthly Trading Volume for Nasdaq Firms

Figure 3 compares monthly trading volume for firms with VC backing against those without VC backing
for Nasdaq firms for the two years following the IPO.

volume and turnover for the NYSE firms classified according to the presence or absence of VC
backing. The non-VC firms that choose the NYSE have higher trading volumes than the VC
firms. Both groups of firms exhibit an increase in volume around month six. Trading volume
for the VC-backed firms increases 35% from month five to month seven, whereas the increase
in volume for the other firms is 13%. Figure 3 shows monthly trading volume and turnover for
Nasdaq firms. The VC-backed IPO firms have higher trading volumes than the non-VC firms in
each month, and they have a marked increase in volume from month five to month seven. Trading
volumes for the VC-backed firms increase by 61% from month five to month seven, compared
to an increase of 22% for non-VC firms. Thus, our data are consistent with the earlier work and
further document that the selling by venture capitalists when lockup agreements expire is more
pronounced for IPO firms trading on Nasdaq.

IV. Logistic Regressions of the Listing Decision

Here we examine the IPO firm’s listing choice and whether it is influenced by either the presence
of venture capitalist backers or by other by pre-IPO investors who intend to sell restricted shares
following the IPO. We estimate a set of logistic regression models that explain IPO firms’ listing
choices, and that include a VC variable and/or a measure of actual or expected restricted stock
Anderson & Dyl • IPO Listings: Where and Why? 37
sales as independent variables. The latter variable is our proxy for the selling intentions of pre-IPO
investors.
The dependent variable in each of the logistic regressions is the choice of the stock market
on which the firm’s stock will trade. The variable is an indicator variable that equals one if the
firm chooses the NYSE, and zero if the firm chooses Nasdaq. Most of the independent variables
in the regression are characteristics of the IPO and of the firm, which we include in the model
to control for their influence on the listing choice. These variables are firm size, underwriter
quality, whether or not the IPO firm is a high-tech firm, and the difference in listing fees between
the NYSE and Nasdaq for the firm. Each of the models also contains an independent variable
that either indicates VC backing or measures the prevalence of restricted stock sales during the
two years following the IPO. The rationale for including each of these independent variables
follows.
Firm Size. Corwin and Harris (2001) show that, among firms that qualify for an NYSE list-
ing, firm size is far and away the most significant determinant of IPO firms’ listing choices.
We incorporate firm size into our analysis by using both the offer price in the IPO and the
number of shares outstanding after the IPO. We include both variables separately because Ta-
ble I shows that in addition to being larger than the Nasdaq firms, the NYSE firms in our
sample have both higher offer prices and more shares outstanding. Fernando, Krishnamurthy,
and Spindt (2004) present evidence that the price levels of IPO shares per se affect IPO
outcomes.
Underwriter Quality. We include this variable to control for the possibility that high-quality
underwriters tend to recommend the NYSE and vice versa. Carter and Manaster (1990) and
Carter, Dark, and Singh (1998) are among several researchers who document the importance of
underwriter reputation as a factor in IPO success. We measure underwriter quality as the market
share ranking of the lead underwriter for the IPO as described earlier.
Listing Fee Difference. Ceteris paribus, it is more costly for a firm to list on the NYSE than
on Nasdaq, which could deter firms from listing on the NYSE. We measure this difference as the
costs of listing on the NYSE minus the costs of listing on Nasdaq. Listing costs include the initial
listing fee and one year of the annual listing fee. Corwin and Harris (2001) include this variable
in their analysis, and report that it was not significant during the period they study.
High-Tech Industry. Firms are classified as high-tech if they operate in the computer, cellular,
healthcare, or data-processing industries. Both Wall Street lore and the descriptive statistics
reported earlier in Table II suggest that high-technology firms tend to gravitate to Nasdaq, so we
include this designation as an independent variable. The variable equals one if the firm’s industry
is classified as high-tech by SDC and zero otherwise.
VC Backing. Because IPOs are an exit strategy for venture capitalists and other private equity
investors, venture capitalists may be motivated to select the stock market that facilitates the sale of
their shares (i.e., Nasdaq). Therefore, we include whether the firm had VC backing as a variable
in the regression. Since there is considerable overlap between firms in high-tech industries and
firms with VC backing, we estimate separate regressions with and without both variables. We
measure VC backing as an indicator variable, equal to one if venture capitalists are pre-IPO
investors in the firm, and zero otherwise.
Restricted Stock Sales. We also include restricted stock sales following the IPO as an indepen-
dent variable in each of the models. Although this selling actually occurs after the listing decision
has been made, presumably investors know their selling intentions at the time they decide to list
the stock on the NYSE or Nasdaq, and this decision could affect the listing choice. This variable
is the number of restricted shares—that is, those subject to the requirements of Rule 144—that
are sold during the two years following the IPO.
38 Financial Management • Spring 2008
Table VI. Logistic Regressions of the Listing Choice of IPO
Firms from 1993 to 2000

The dependent variable is the stock market listing for 640 firms that go public during 1993-2000 (NYSE =
1, Nasdaq = 0). The independent variables are the offer price of the IPO, the log of the number of shares
outstanding, underwriter quality measured as the market share of the lead underwriter, the difference in
listing fees for the firm (NYSE fees minus Nasdaq fees), whether the firm is in a high technology industry,
the percentage of shares owned by CEO and chairman prior to the IPO, whether or not the firm had venture
capital backing, actual sales of restricted shares during the two years following the IPO, and expected sales
of restricted shares during the two years following the IPO.

Independent Variable Model 1 Model 2 Model 3 Model 4 Model 5 Model 6

Intercept −8.3473∗∗∗ −8.7027∗∗∗ −8.5117∗∗∗ −8.3206∗∗∗ −9.9555∗∗∗ −9.9329∗∗∗
Offer price ($) 0.0861∗∗∗ 0.0748∗∗ 0.0787∗∗ 0.0890∗∗∗ 0.1197∗∗ 0.1194∗∗
Shares outstanding (log) 1.0840∗∗∗ 1.2365∗∗∗ 1.2005∗∗∗ 0.9579∗∗∗ 1.7400∗∗∗ 1.7330∗∗∗
∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗
Underwriter quality 0.5957 0.5829 0.5811 0.5704 0.5434 0.5438∗∗∗
Listing fee difference −0.0332 1.9949 1.6808 2.8044 −1.4377 −1.4281
($ millions)
High-tech industry −1.6835∗∗∗ −1.5934∗∗∗ −1.5042∗∗∗ – −1.3902∗∗∗ −1.3890∗∗∗
(yes = 1, no = 0)
VC backing (yes = 1, −0.8874∗∗∗ – −0.6439∗∗ −0.8736∗∗∗ – −0.3063
no = 0)
Actual sales of restricted – −0.5547∗∗∗ −0.5185∗∗∗ −0.6096∗∗∗ – –
stock
Expected sales of – – – – −1.2657∗∗∗ −1.2537∗∗∗
restricted stock
Pseudo R squared 31.88% 34.23% 34.94% 30.68% 35.89% 35.89%
∗∗∗
Significant at the 0.01 level.
∗∗
Significant at the 0.05 level.

We report the results of the logistic analyses of the listing decision in Table VI. The dependent
variable in these logit models equals one if the firm chooses to list on the NYSE and zero
if the firm chooses Nasdaq. We report the results for various combinations of independent
variables, but most of the models have five variables in common. The sign and significance of the
coefficients on these variables is essentially the same for each model, so we discuss them as a single
finding.
Consistent with both Corwin and Harris (2001) and the univariate statistics in Table II, larger
firms are more likely to select the NYSE. The coefficients on both of the size measures—offer
price and shares outstanding—are positive. The former is generally significant at the 5% level
and the latter is significant at the 1% level. We note that the results are similar when we use
a single measure of firm size—market capitalization—in place of two separate variables. The
coefficient on size is significant at the 1% level, but the pseudo R squared of each regression is
slightly lower.
The coefficient on underwriter quality is positive and significant at the 1% level in each model.
IPO firms with more prestigious lead underwriters are more likely to list on the NYSE and vice
versa. Either more prestigious underwriters recommend that their clients list on the NYSE or
else larger, better established firms have more prestigious underwriters and choose the more
prestigious trading location for their stock.
The difference in listing fees between listing on the NYSE and listing on Nasdaq has the
expected negative sign, but it is not significant. In view of the magnitude of the difference in listing
Anderson & Dyl • IPO Listings: Where and Why? 39
fees, this result is not surprising. Table I shows that the differences are small relative to the size of
the IPOs in our sample. This result is consistent with the findings of Corwin and Harris (2001).
The high-tech industry variable is negative and significant at the 1% level. Firms in high-tech
industries are more likely to list on Nasdaq. Corwin and Harris (2001) report a similar result
without comment, although they note that IPO firms tend to list on the same exchange as their
industry peers. 3 The predilection of technology firms for the Nasdaq stock market is probably
an historical artifact. In the 1970s and 1980s, many IPOs of technology firms that subsequently
became extraordinarily successful, such as Intel and Microsoft, traded on Nasdaq. These firms
later found no compelling reason to switch to the NYSE, and Nasdaq has thus become the stock
market in which technology firms are expected to trade.
In addition to the five variables discussed earlier, Model 1 also contains a dummy variable
(VC Backing) denoting that the IPO firm has financing from venture capitalists. The coefficient
on VC Backing is negative and significant, indicating that firms backed by venture capitalists
are more likely to choose Nasdaq. Venture capitalists have an incentive to influence firms in
this direction, since the selling restrictions under Rule 144 apply to the volume of sales by
individuals, and venture capitalists’ sales of restricted shares tend to be larger than those of other
pre-IPO investors. Cao, Field, and Hanka (2004) report that that the value (number) of insider
sales following lockup expirations is distributed as follows: 33% (8%) by venture capitalists, 35%
(68%) by employees, and 31% (24%) by other pre-IPO shareholders. These numbers suggest that
venture capitalists’ individual stock sales are eight times larger than per capita stock sales by
employees and more than three times larger than per capita sales by other pre-IPO investors, so
venture capitalists have a powerful incentive to attempt to mitigate the Rule 144’s limits on the
sales of restricted shares. Moreover, venture capitalists frequently serve on the boards of directors
of the firms they finance and are otherwise influential in financial decisions; they are clearly in
a position to affect the firm’s choice of trading location for its stock.
In Model 2, we replace the VC variable with a measure of restricted stock sales during the two
years following the IPO. We use this variable as an indicator of pre-IPO selling intentions. Initially,
we consider the effect of VC and restricted stock sales variable separately, because the regression
models in Table V show that VC backing is a major determinant of restricted stock sales. The
coefficient on restricted stock sales also has a negative sign and is significant, indicating that
firms in which the pre-IPO investors intend to sell more of their shares after the IPO are less
likely to list on the NYSE. This result is consistent with IPO firms choosing Nasdaq to mitigate
Rule 144’s limits on restricted stock. Moreover, the R squared of 34.23% for Model 2 compared
to an R squared of 31.88% for Model 1, shows the restricted stock sale variable does more to
explain the listing choice that the presence of the venture capitalist alone.
In Model 3, we investigate whether the presence of venture capitalists influences the listing
choice over and above their effect on restricted stock sales. Table V shows the relation between
venture capitalist backing and sales of restricted stock. Model 3 expands Model 2 by adding the
variable denoting VC-backing. This variable has a negative sign and is significant, albeit at the
5% level instead of the 1% level, and the R squared of the regression increases from 34.23% for
Model 2 to 34.94% for Model 3. Venture capitalists have an effect on the listing choice over and
above their selling behavior, which is represented in the Actual Sales of Restricted Stock variable.
The rationale for including a variable denoting the high-tech industry classification is arguably
anecdotal and it is not entirely clear what this variable is capturing. We noted earlier that venture
capitalists are more plentiful in high-tech industries than in the overall sample of IPO firms. To

3
Adding a complete set of industry dummy variables to our logistic regressions does not improve the explanatory power
of the models.
40 Financial Management • Spring 2008
investigate this issue further, we estimate the regression without the high-tech industry variable.
Model 4, Table V, reports the results. The only noteworthy change from Model 3 is that the R
squared of the regression drops from 34.94% to 30.68%. Apparently, firms in high-tech industries
prefer to trade on Nasdaq, rather than the NYSE, regardless of the presence of venture capitalists
and/or the owners, post-IPO selling intentions.
A potential problem with the analysis of the listing decision reported as Models 2 and 3 in
Table VI is the interpretation of the variable that measures post-IPO sales of restricted stock. This
variable reports selling behavior that takes place after the listing decision is made and the stock
is actually trading in the marketplace. Although these ex post measures of restricted stock sales
may be good proxies for investors’ ex ante intentions, actual selling behavior also depends on the
outcome of the listing decision. That is, the Rule 144 limits on sales of restricted shares constrain
the selling behavior of pre-IPO investors more when a firm chooses to list on the NYSE than when
it chooses Nasdaq (see Model 4 in Table V). To formally test for the presence of endogeneity,
we use the parameters of the regression equation shown as Model 1 in Table V to calculate the
expected values of the dependent variable for each firm, and we include the residuals from these
models as independent variables in the logistic regressions. We then reestimate the parameters of
the logistic regressions and find that the coefficient on the residual of this regression is significant
in each model, indicating endogeneity (see Wooldridge, 2002).
To address this issue, in Table VI we use Model 1 from Table V to replace the restricted stock
sales variables in logistic Models 2 and 3 with the expected values for sales of restricted stock
estimated using Model 1 from Table V. Models 5 and 6 in Table VI report the results. Using
this instrumental variable in place of the endogenous measure of restricted stock sales does not
alter our findings in Model 2. The signs and significance of the coefficients on the independent
variables in Models 5 are essentially identical to those in Models 2, and the pseudo R squared of the
model increases to 35.89%. In Model 6, which is the replication of Model 3 using the instrumental
variable for restricted stock sales in place of actual sales, the coefficient on VC backing is no
longer significant. This result is simply an artifact of the instrumental variable. The regression
we use to estimate the instrumental variable—Expected Sales of Restricted Stock—contains the
same dummy variable denoting VC backing that is also included in Model 6.

V. Conclusion
The underlying rationale for the SEC’s Rule 144 is to enable individuals to conduct day-to-
day trading transactions involving unregistered stock, which are permitted under Section 4(1)
of the Securities Act of 1933, but to prohibit firms and/or investors from making secondary
distributions of large amounts of unregistered stock in violation of the Act. The rule limits the
number of restricted shares that an individual can sell in the open market during any three-month
period to either 1% of shares outstanding or the average weekly trading volume reported during
the preceding four weeks. Because of the different protocols for reporting trading volume on the
NYSE compared to Nasdaq, this provision of Rule 144 affects the listing decisions of firms that
go public. It is the reason why many IPO firms that are eligible for listing on the NYSE instead
choose to have their shares traded on Nasdaq. From 1993 through 2000, 35% (224) of the 640
IPOs eligible to list on the NYSE chose to trade on Nasdaq.
We examine open market sales of restricted shares for a sample of firms that go public between
1993 and 2000 and find that Nasdaq firms sell more restricted stock following the IPO than
do the NYSE firms. Of the firms that chose Nasdaq, 64% had sales of restricted stock in the
two years following the IPO compared to only 40% of the NYSE firms. Tobit models show that
Anderson & Dyl • IPO Listings: Where and Why? 41
restricted stock sales are positively related to the stock’s volatility and to the presence of venture
capitalists as pre-IPO investors, are unrelated to whether or not the firm went public during a hot
IPO market, and are negatively related to either whether or not the firm goes public before 1998
or chooses to list on the NYSE.
After controlling for firm size, underwriter quality, and other variables, we find that the listing
decision is significantly related to VC backing and to post-IPO restricted stock selling. Both the
presence of venture capitalists and the number of restricted shares sold are strongly related to the
listing decision in a manner consistent with firms choosing Nasdaq over the NYSE to mitigate
the effect of the Rule 144’s limits on open market sales of restricted shares.
When we replicate the analyses using instrumental variables to estimate expected restricted
stock sales, we reach the same conclusions. Overall, firms appear to list on Nasdaq to obtain
advantageous regulatory treatment under SEC Rule 144. It is not clear if the advantage they
obtain is sufficient to be considered a de facto circumvention of the registration requirements
of the Securities Act of 1933. Nonetheless, one implication of our findings is that Rule 144 has
not been enforced even-handedly across markets. The SEC should consider either change Rule
144 or adopting a more consistent measure of trading volume across auction and dealer markets
to eliminate this discrepancy, so that companies will receive equal treatment under the securities
laws regardless of the trading venue they choose for their stock.

Appendix. Summary of Requirements for Listing on the NYSE

Years Requirements
Panel A. Size Requirements
1993-1994 1. Net tangible assets greater than or equal to $18 million
2. IPO proceeds greater than or equal to $18 million
3. Number of public shares greater than or equal to 1.1 million
1995-1997 1. Net tangible assets greater than or equal to $40 million
2. IPO proceeds greater than or equal to $40 million
3. Number of public shares greater than or equal to 1.1 million
1998 1. IPO proceeds greater than or equal to $40 million
2. Shares offered greater than or equal to 1.1 million
1999-2000 1. IPO proceeds greater than or equal to $60 million
2. Number of public shares greater than or equal to 1.1 million

Panel B. Earnings Requirements

All years 1. Earnings in current year greater than or equal to 2.5 million and
earnings data for previous two years
2. Earnings in current year greater than or equal to 2.5 million,
and year t−1 and year t−2 earnings greater than 2 million OR
total earnings for current, t−1, and t−2 greater than or equal to
6.5 million and current earnings greater than or equal to 4.5
million and earnings in year t−1 and t−2 greater than 0

References

Anderson, A.M., E.A. Dyl, and L. Krigman, 2005, “IPO Listing Choices and the Regulation of Post-IPO
Sales of Restricted Stock,” Lehigh University, University of Arizona, and Babson College Working Paper.
42 Financial Management • Spring 2008
Ang, J.S. and J.C. Brau, 2003, “Concealing and Confounding Adverse Signals: Insider Wealth-Maximizing
Behavior in the IPO Process,” Journal of Financial Economics 67, 149-172.
Black, B. and R. Gilson, 1998, “Venture Capital and the Structure of Capital Markets: Bank versus Stock
Markets,” Journal of Financial Economics 47, 243-277.
Bradley, D.J., B.D. Jordan, H. Yi, and I.C. Roten, 2001, “Venture Capital and IPO Lockup Expiration: An
Empirical Analysis,” Journal of Financial Research 24, 465-493.
Cao, C., L.C. Field, and G. Hanka, 2004, “Does Insider Trading Impair Market Liquidity? Evidence from
IPO Lockup Expirations,” Journal of Financial and Quantitative Analysis 39, 25-46.
Carter, R., F. Dark, and A. Singh, 1998, “Underwriter Reputation, Initial Returns and Long-Run Performance
of IPO Stocks,” Journal of Finance 53, 285-311.
Carter, R. and S. Manaster, 1990, “Initial Public Offerings and Underwriter Reputation,” Journal of Finance
45, 1045-1067.
Corwin, S.A. and J.H. Harris, 2001, “The Initial Listing Decisions of Firms that Go Public,” Financial
Management 30, 35-55.
Cowan, A.R., R.B. Carter, F.H. Dark, and A.K. Singh, 1992, “Explaining the Listing Choices of Nasdaq
Firms,” Financial Management 21, 73-86.
Fernando, C.S., S. Krishnamurthy, and P.A. Spindt, 2004, “Are Share Price Levels Informative? Evidence
from the Ownership, Pricing, Turnover and Performance of IPO Firms,” Journal of Financial Markets 7,
377-403.
Field, L.C. and G. Hanka, 2001, “The Expiration of IPO Share Lock-Ups,” Journal of Finance 56, 471-500.
Goergen, M., L. Renneboog, and A. Khurshed, 2006, “Explaining the Diversity in Shareholder Lockup
Agreements,” Journal of Financial Intermediation 15, 254-280.
Hall, B. and K. Murphy, 2002, “Stock Options for Undiversified Executives,” Journal of Accounting and
Economics 33, 3-42.
Helwege, J. and N. Liang, 2004, “Initial Public Offerings in Hot and Cold Markets,” Journal of Financial
and Quantitative Analysis 39, 541-569.
Helwege, J. and F. Packer, 2005, “Private Matters,” University of Arizona and Bank for International
Settlements Working Paper.
Jain, P.K. and J. Kim, 2006, “Investor Recognition, Liquidity, and Exchange Listings in the Reformed
Markets,” Financial Management 35, 21-42.
Kadlec, G.B. and J.J. McConnell, 1994, “The Effect of Market Segmentation and Illiquidity on Asset Prices:
Evidence from Exchange Listings,” Journal of Finance 49, 611-636.
Kahl, M., J. Liu, and F.A. Longstaff, 2003, “Paper Millionaires: How Valuable is Stock to a Stockholder
Who is Restricted from Selling It?” Journal of Financial Economics 67, 385-410.
Leland, H. and D. Pyle, 1977, “Informational Asymmetries, Financial Structure, and Financial Intermedia-
tion,” Journal of Finance 32, 371-387.
Loughran, T. and J. Ritter, 2004, “Why has IPO Underpricing Changed Over Time?” Financial Management
33, 5-37.
Lowry, M. and G.W. Schwert, 2002, “IPO Market Cycles: Bubbles or Sequential Learning?” Journal of
Finance 57, 1171-1200.
Anderson & Dyl • IPO Listings: Where and Why? 43
Meulbroek, L., 2001, “The Efficiency of Equity-Linked Compensation: Understanding the Full Cost of
Awarding Executive Stock Options,” Financial Management 30, 5-30.
Steinberg, M.I. and J.P. Kempler, 1988-1989, “The Application and Effectiveness of SEC Rule 144,” Ohio
State Law Journal 49, 473-500.
Wooldridge, J.M., 2002, Econometric Analysis of Cross Section and Panel Data, Cambridge, MA,The MIT
Press.
44 Financial Management • Spring 2008