ADRs, Analysts, and Accuracy: Does Cross Listing in the U.S. Improve a Firm's Information Environment and Increase Market Value?

ADRs, Analysts, and Accuracy: Does Cross Listing in the U.S.
Improve a Firm's Information Environment and Increase Market
Value?
Mark H. Lang, • Karl V. Lins,† and Darius P. Miller‡
Forthcoming in The Journal of Accounting Research, 2003
Abstract
This paper investigates the relation between cross listing in the U.S. and the information
environment of non-U.S. firms. We find that firms that cross list on U.S. exchanges have greater
analyst coverage and increased forecast accuracy relative to firms that are not cross listed. A
time-series analysis shows that a change in analyst coverage and forecast accuracy occurs around
cross listing. We also document that firms that have more analyst coverage and higher forecast
accuracy have higher valuations. Further, the change in firm value around cross listing is
correlated with changes in analyst following and forecast accuracy, suggesting that cross listing
enhances firm value through its effect on the firm’s information environment. Our findings
support the hypothesis that cross-listed firms have better information environments, which are
associated with higher market valuations.

• The University of North Carolina; †The University of Utah; ‡Indiana University. We thank Ray Ball (the editor)
and Christian Leuz (the referee) for detailed suggestions on improving the paper. We also thank Walt Blacconiere,
Mike Lemmon, Jim Linck, Laureen Maines, Marlene Plumlee, Lynn Rees, Henri Servaes, Charles Trzcinka and
seminar participants at the 2002 Journal of Accounting Research Conference, Indiana University, the Eighth Annual
Georgia Tech/Fortis International Finance Conference, and the Darden Valuation in Emerging Markets conference
for helpful comments. Alan Montgomery provided valuable research assistance. We are grateful for the analyst
forecast information provided by I/B/E/S Inc. Email addresses for the authors are langm@bschool.unc.edu,
finkvl@business.utah.edu, and damiller@indiana.edu.

---

1. Introduction
In this paper, we examine cross listing in the U.S. and its relation to the information
environment of non-U.S. firms.1 A large literature on international cross listing suggests that
information considerations are a key factor in cross listing decisions. However, there is little
direct empirical evidence on the relation between cross listing and the information environment
of the firm. We document several empirical findings on this subject. First, we show cross-
sectionally that non-U.S. firms that are listed on U.S. exchanges have greater analyst coverage
and increased forecast accuracy relative to other non-U.S. firms. Second, we perform time-
series analyses and find that a change in analyst coverage and forecast accuracy occurs around
cross listing. Third, we document that analyst coverage and forecast accuracy are both positively
related to firm value. Finally, we show that the change in value around cross listing is correlated
with changes in analyst following and forecast accuracy, suggesting that cross listing increases
firm value through its effect on the firm’s information environment. Our results are robust to
adjustments for the potential endogeneity of the listing decision and simultaneity between
analyst following and forecast accuracy. Overall, our findings support the hypothesis that
important informational effects occur with cross listing and that these effects are positively
associated with firm value.
Foerster and Karolyi [1999] and Miller [1999] document positive average abnormal
announcement returns for non-U.S. firms that issue exchange-listed American Depositary
Receipts (ADRs), while Foerster and Karolyi [2000] document positive long-horizon returns for
such firms that raise capital. Similarly, Errunza and Miller [2000] find a substantial decline in a

1 Our notion of information environment is similar to the concept of “corporate transparency” in Bushman, Piotroski
and Smith [2001], broadly construed to include the effects of “corporate reporting, private information acquisition
and information dissemination.”
1

---

firm’s cost of capital after an ADR. These papers and others offer a number of explanations for
why cross listing on a U.S. stock exchange adds value.2 However, a crucial component in
almost all of these explanations is the firm’s information environment. The notion that the
information environment should be a function of cross listing is natural, since, as discussed in
Coffee [2002], cross listing firms subject themselves to (1) increased enforcement by the SEC,
(2) a more demanding litigation environment and (3) enhanced disclosure and reconciliation to
U.S. GAAP. In addition, cross listing firms may face more scrutiny from investors, more
pressure to provide guidance than they did in their home markets, and increased scrutiny from
their auditors. Firms that list in U.S. markets are, in effect, “bonding” themselves to an
increased level of disclosure and scrutiny. These changes in transparency could affect firm value
by decreasing the cost of capital, increasing the cash flows that ultimately accrue to
shareholders, or both.
For the cost of capital, Merton’s [1987] investor recognition hypothesis is often used to
argue that a U.S. listing creates value because the enhanced disclosure environment reduces the
cost of following the firm. This increases the investor base and, therefore, the demand for the
firm’s securities. Barry and Brown [1985] suggest that cost of capital is a function of
“estimation risk” and the better investors are able to assess the prospects for a company, the
lower its expected cost of capital. Along this line of reasoning, Lang and Lundholm [1996]
show that analysts’ forecasts are more accurate for firms that disclose more, while Gebhardt,
Lee, and Swaminathan [2001] find that firms with more accurate forecasts enjoy a lower implied
cost of capital. Thus, if investors are able to more accurately assess the prospects for a cross-
listed firm, its cost of capital should be reduced. Taken together, the investor recognition and

2 For references on earlier studies such as Errunza and Losq [1985] and Alexander, Eun, and
2

---

estimation risk hypotheses suggest that a firm’s information environment can play an important
role in determining its cost of capital.
The literature also suggests that the enhanced transparency associated with cross listing
may influence value through pure cash flow effects by reducing agency costs. For example,
cross listing may be associated with improved firm-level corporate governance because it bonds
a firm to greater transparency, which should reduce the potential diversion of a firm’s cash flows
to managers and controlling shareholders (Coffee [1999] and Stulz [1999]). Consistent with this
hypothesis, recent research by Doidge, Karolyi and Stulz [2002] documents that a U.S. listing
creates the most value for firms with higher growth opportunities located in countries that have
poor disclosure and investor protections. Lins, Strickland, and Zenner [2002], Reese and
Weisbach [2002], and Pagano, Roell, and Zechner [2002] argue that cross-listing adds value
because the greater transparency increases the willingness of both international and local
investors to commit capital.
Diamond and Verrecchia [1991] and Leuz and Verrecchia [2000] emphasize the
importance of precommitment in the relation between disclosure and cost of capital. Cross
listing provides a credible commitment to increased disclosure because a firm is subject to
greater regulatory and investor scrutiny, disclosure requirements, and potential legal exposure.
This differentiates cross listing from simply announcing an intention to increase disclosure in the
home market because a cross listed firm cannot easily renege on its commitment if it later turns
out to have bad news which it would prefer not to disclose. It is relatively costly from a
reputational standpoint to delist, since the firm would risk alienating its international investor
base. A similar argument applies to the effect of cross listing on agency problems and cash

Janakiramanan [1987, 1988], see the survey by Karolyi [1998].
3

---

flows; investors are likely to anticipate better incentive alignment for firms that bond themselves
to better investor protection by listing in the U.S.
Despite its theoretical importance, surprisingly little direct evidence on the relation
between a firm’s information environment and cross listing exists.3 One factor that makes
testing this relation difficult is that it is not possible to directly measure a firm’s information
environment. For instance, Bailey, Karolyi, and Salva [2002] use price volatility and volume
reaction to earnings announcements to conduct tests on the information environment of cross-
listed firms. Baker, Nofsinger, and Weaver [2002] test whether “visibility” increases around
cross listing with measures of analyst following and media coverage. Our approach is to follow
prior research by Lang and Lundholm [1996], Healy, Hutton and Palepu [1999], and Gebhardt,
Lee, and Swaminathan [2001] and use the characteristics of analyst forecasts as a proxy for the
information environment. In particular, we focus on two measures: the number of analysts
following the firm and the accuracy of analyst forecasts. Previous studies suggest that having
more analysts with more accurate forecasts indicates a firm with a better information
environment.4

3 There is a substantial body of research examining the nature of reconciling items for cross listed firms as
reported on Form 20-F (summarized in Pownall and Schipper [1999]). Much of that literature examines the
association between reconciling items and share prices to infer whether the information in reconciliations is
"value relevant" and generally finds a significant association. However, evidence on whether the Form 20-
F is the source of the information is mixed since the information release date is generally not clear. Our
interest is different in that we are not concerned about whether specific reconciling items are associated
with share price but, rather, with the question of whether cross listed firms are characterized by generally
richer information environments.
4 Following the prior literature, we view the analyst variables as indicative of, but not necessarily the cause of,
changes in a firm’s information environment. For example, analyst forecast accuracy is intended as a measure of
how well the market understands the firm’s economics. This may partially be a result of analyst activity, but it may
also reflect disclosure by the firm or information gathering by other investors. Similarly, analyst following is
intended to proxy for private information acquisition activities. While the analysts’ research may indeed enhance the
information environment, the same incentives that attract sell-side analysts might also attract buy-side analysts and
other investors. Further, our two measures are not intended as mutually exclusive. For example, forecasts for a
given firm may be more accurate because there are more analysts following the firm. Similarly, we do not believe
the two measures are entirely redundant; forecasts may be accurate because a firm discloses more without more
analysts following. Our empirical tests take these two measures into account both separately and simultaneously.
4

---

Analyst following should be related to cross listing and value for several reasons. First,
to the extent that cross listing increases the quantity of information available to the market, either
because of explicit disclosure requirements or implicit pressure to provide additional information
to analysts and investors, it should reduce the cost of following a firm, which could lead to
increased coverage by investment analysts. Second, cross listing widens the potential investment
base of a firm. This should be associated with increased analyst activity because analysts are
likely to focus on firms that investors find interesting and investors are more likely to consider
firms followed by analysts. Similarly, analysts may be attracted to cross listing firms because
they perceive them to be of higher quality and therefore of more potential interest to their
potential investor base. These predictions are also consistent with the investor relations
literature, which suggests that a benefit of increased disclosure is increased analyst following.5
Finally, additional analyst following should also bring about more scrutiny, which, in the
presence of agency costs, should improve firm value by increasing the cash flows that accrue to
shareholders (Lang, Lins, and Miller [2002]). In summary, increased analyst following around
cross-listing can influence aspects of the firm’s information environment that have been argued
to affect both cash flows and the cost of capital.
The above arguments suggest that a firm’s disclosures and the information produced by
analysts complement each other. Consistent with this hypothesis, Lang and Lundholm [1996]
and Healy, Hutton, and Palepu [1999] find that, for U.S. firms, increased disclosure is associated
with higher analyst following. However, it is important to note that while empirical support
exists for a positive relation between analyst activity and increased disclosure, the direction of
the association is not obvious. For instance, Lang and Lundholm [1996] note that to the extent

5 We focus our discussion on sell-side analysts since the IBES data are primarily based on their forecasts.
5

---

that extra disclosure levels the playing field among analysts, it could reduce any one analyst’s
competitive advantage, which would lessen incentives to cover the firm. Botosan [1997] finds
that when firms already have a high analyst following, increased disclosure is not associated with
a reduction in the cost of capital, suggesting that analysts and disclosure function as substitutes.6
These latter two papers indicate that additional public disclosure could drive out private
information acquisition, resulting in an ambiguous effect on total information in the market.
In an international context, Bushman, Piotroski, and Smith [2001, 2002] provide
evidence of a positive correlation between analyst following and disclosure and investor
protection, suggesting that analysts might be attracted to firms that cross list on U.S. markets.
On the other hand, Chang, Khanna, and Palepu [2000] find that companies in code law countries
tend to have greater analyst following than companies in common law countries. This finding
suggests that by subjecting itself to U.S. regulatory requirements, a cross-listed firm might
discourage analyst following. As a result, the direction of the predicted relation between cross
listing and analyst following is an empirical issue.
Similarly, the link between analyst following and value is not necessarily positive. For
example, if analysts primarily gather private information, their activities could actually increase
cost of capital by raising transactions costs and discouraging uninformed investors from
purchasing shares (see, for example, Diamond and Verrecchia [1991]). Although such an effect
on valuation might be offset by an increase in investor interest, reduced uncertainty, and reduced
agency conflicts within cross listed firms, the expected relation between analyst following and
valuation on net is not clear ex ante.

6 However, using a longer time period and a larger sample of firms, Botosan and Plumlee [2002] document a
negative association between annual report disclosure and cost of capital for highly followed firms.
6

---

We also examine the ex post accuracy of analyst forecasts. To the extent that cross
listing directly or indirectly increases the amount of information available about the firm, one
can argue that it will improve the accuracy of analyst forecasts. Improved accuracy should
reduce the cost of capital through its effect on estimation risk. As noted earlier, Gebhardt, Lee,
and Swaminathan [2001] find that firms with lower forecast errors have lower implied costs of
capital. Improved accuracy may also lessen agency problems to the extent that managers are
held more accountable for the details of their firm’s cash flows.
However, as with analyst following, the direction of the association between cross listing
and forecast accuracy is not obvious ex ante. While cross-listed firms are not required to change
local GAAP reporting when they list on U.S. markets, Lang, Raedy, and Yetman [2001] suggest
that firms change in ways that make earnings more volatile and more similar to U.S. firms.7
Similarly, non-U.S. analysts might find it harder to predict earnings following a movement
toward U.S. GAAP because of greater familiarity with the local GAAP. To the extent that
earnings become less predictable around cross listing, it should bias against our finding results.
We explicitly include controls for earnings surprise in our analysis which should mitigate the
effects of increased volatility.
In summary, while much of the literature suggests that cross listed firms should have a
richer information environment in terms of greater analyst following and forecast accuracy, there
are also reasons to believe that the relation between cross listing and analyst following and
forecast accuracy might be negative, leaving the issue as an empirical question. This paper adds
to the literature by examining analyst activity around cross listing to see if, in fact, cross-listing is

7 Ball [2001] discusses the evidence on cross-country differences in accounting and concludes that environments
like the U.S. focus on the timely recognition of losses (rather than smoothing them over time), which is consistent
with an increase in earnings volatility after cross listing. Similarly, Bailey, Karolyi, and Salva [2002] find that stock
7

---

associated with greater analyst following and forecast accuracy. We then investigate whether
these changes in the information environment are linked to firm value.
The rest of the paper is organized as follows: Section 2 describes the data. Section 3
examines the impact of cross listing on analyst following and forecast accuracy. Section 4
details the impact of analyst following and performance on firm value. Section 5 offers
concluding remarks.
2. Data
To examine the firm’s information environment, we focus much of our analysis on the
levels of our information variables and consider our changes analysis as supplementary. We
take this approach for several reasons. First, the predictions from the prior literature for the level
of the information environment following cross listing are clearest. In particular, depending on
the view of cross listing, it is possible to envision situations in which the information
environment is important, but is not necessarily reflected in changes around cross listing. For
example, Coffee [1999] and Leuz and Verrecchia [2000] argue that the important aspect of cross
listing from a valuation perspective is the commitment to increased disclosure, rather than the
increase in disclosure itself. Cantale [1998], Fuerst [1998], Moel [1999], and Huddart, Hughes,
and Brunnermeier [1999] argue that a firm that has history of transparency will still have an
incentive to list because it signals its commitment to continuing that policy, even when it faces
circumstances under which it might wish not to disclose. Following this reasoning, in Doidge et
al. [2002] the decision to cross list indicates a firm with well-aligned incentives that is therefore

return and volume reactions to earnings announcements typically increase once a stock cross-lists in the U.S.,
suggesting that earnings becomes less predictable and more informative.
8

---

willing to submit itself to scrutiny. It is important to note, however, that the firm may have also
had a rich information environment prior to the cross listing.
Closely related, even if the information environment explicitly changes because of the
cross listing, the timing may not be clear. For example, it is unlikely that a firm anticipating a
U.S. listing would increase disclosure suddenly following the listing or that analysts would
suddenly increase their activity. Rather, as argued by Bradshaw and Miller [2002] and Lang,
Raedy and Yetman [2002], it is likely that firms increase disclosure gradually in anticipation of
the listing by moving their accounting closer to U.S. GAAP, increasing their footnote disclosure,
and communicating more freely with analysts and investors. As a result, it is difficult to know
the window over which to compute the change. We consider long windows to increase the
probability that we capture the entire effect, recognizing that longer windows increase the
probability of confounding effects. Finally, evaluating changes substantially reduces both our
sample size and our ability to include historical control variables since our data sources begin
large-scale coverage only in the early to mid 1990s, and most cross listed firms became listed
prior to the mid 1990s.
To examine the extent and accuracy of analyst activity, we utilize data from the
Historical I/B/E/S International database. The main results of the paper are for the year 1996. We
choose this year because the year-to-year increase in I/B/E/S coverage of firms begins to slow
substantially after 1996 and we want to capture any benefits of ADR listings as far back in time
as possible. We use data from the eleventh month of the fiscal year to calculate the number of
analysts following a company and the forecast accuracy, as O’Brien and Bhushan [1990]
9

---