Does Cross-Listing Mitigate Insider Trading?

1
Does Cross-Listing Mitigate Insider Trading?

• Adriana Korczak and Meziane Lasfer
• Cass Business School, London

2
Introduction

• Evidence that insiders trade profitably around
  major corporate events using private information
• Bankruptcy protection Seyhun Bradley (1997)
• New issues Karpoff and Lee (1991)
• Buybacks Lee, Mikkelson Partch (1992)
• Earnings forecasts Penman (1982)
• Takeovers Seyhun (1990), Bris (2005)
• Dividend announcements John and Lang (1991)
• Exchange listings/de-listings Lamba and Khan
  (1999)
• Evidence that insiders earn significant
  exceptional returns
• US Jaffe (1974), Finnerty (1976), Seyhun (1986),
  Lakonishok and Lee (2001)
• U.K Pope et al (1990), Gregory et al (1994)
  in other countries
• But is Insider Trading profitable after
  transaction costs?

3
Issues Should insider trading be regulated?

• What and How to regulate Controversies
• What is insider trading and who is the insider
• How to treat non-information trading (e.g.,
  portfolio changes, liquidity) and trading on
  miss-valuation
• Insider trading should not be regulated because
• It increases market efficiency, thus,
• Prices will reflect all information Closer to
  strong form EMH
• Signalling Buy (sell) trades to signal under-
  (over-) valuation
• Insider trading should be regulated because
• Trading on private information implies transfer
  of wealth
• Decrease market efficiency through
• Reduction in liquidity
• Informed investors set up strategies to mimic
  insider trades

4
Objective of the paper

• Test whether cross-listing mitigates the trading
  on insider information
• The legal and reputational bonding hypotheses
• UK and US roughly same governance, thus not
  testing the bonding hypothesis as defined by
  (Cofee 1999, 2002 Stulz, 1999)
• Cross-listed companies are subject to both
  domestic and foreign Legislation
• US and UK are relatively complementary Table 1
• Increased disclosure requirements
• Less information asymmetries because more
  thorough investor monitoring
• Stronger bad image effects

5
Cross-listing

• Parallel listing on domestic and foreign stock
  exchanges
• Particularly popular and widely investigated over
  the last 15-20 years

Listing venue Disclosure requirements Capitalraising
Rule 144A Portal Minimal compliance yes
ADR Level I OTC Partial compliance no
ADR Level II AMEX / NASDAQ / NYSE Full compliance no
ADR Level III AMEX / NASDAQ / NYSE Full compliance yes
6
Data

• Source
• Insider trading - Director Deals Ltd.
• Cross-listing - BoNY, NASDAQ/NYSE/AMEX
• Stock prices, accounting data and news - Perfect
  Analysis
• Sample
• 1999-2003
• 928 UK companies (CL 115, 12)
• Total number of observations - NALL13,529 (CL
  18, BuyALL 78 (CL DL))

7
Description of the data (1) Table 2
Cross-Listed Cos (CL) Cross-Listed Cos (CL) Cross-Listed Cos (CL) Domestically-Listed Companies (DL) Domestically-Listed Companies (DL) Domestically-Listed Companies (DL) t CL DL t CL DL Mann-Whitney
Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years) Panel B Fundamentals (Firm-Years)
Buy Trades Market Cap (m) Dividend Yield M/B ROA Buy Trades Market Cap (m) Dividend Yield M/B ROA 19,512 5.00 7.78 0.03 4,845 3.91 1.99 0.04 871 5.17 2.31 0.02 143 4.28 1.41 0.01 18.77 -1.10 2.18 4.71 18.77 -1.10 2.18 4.71 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Sell Trades Market Cap (m) Dividend Yield M/B ROA Sell Trades Market Cap (m) Dividend Yield M/B ROA 18,011 2.97 19.56 0.07 5,642 2.45 3.07 0.07 639 3.22 3.25 0.03 170 2.40 2.25 0.02 10.27 -0.94 1.78 11.41 10.27 -0.94 1.78 11.41 0.00 0.36 0.00 0.00 0.00 0.36 0.00 0.00
Mean Median Mean Median
8
Description of the data (2) Table 2
CL DL Mean
Median Mean Median t
MW
9
Methodology

• Event study methodology
• Event day day 0
• Insider trading announcement date
• Insider trading date
• Event window -100 100
• Estimation window -360 -101
• News announcements
• Regressions
• OLS
• To account for fundamental characteristics of
  cross-listed firms (Reese and Weisbach, 2002
  Doidge et al., 2004) Larger, higher growth and
  profitability
• 2SLS and 2-stage Heckman estimation (Heckman,
  1978)

10
Summary of the results
Sell Trades
CL
DL
DL
CL
Buy Trades
11
Empirical Results
12
OLS Regressions
13
Regressions Selectivity Bias
14
Robustness checks

• Confounding events -5, 5
• Same results
• Announcement day vs. Trading day
• Similar results
• Announcement dates provide more information than
  trading dates
• Bull vs. bear markets
• Cross-listed companies More information in bear
  period
• More differences in domestically-listed firms
• Alternative event study methodologies
• Same results using market adjusted model, mean
  adjusted model
• Control sample Size effect, similar results

15
Impact of News AnnouncementsPre-event Buy
trades
16
Impact of news announcementsPost-event Buy
trades
17
Impact of news Pre-Sell trades
18
Impact of news Post-Sell Trades
19
(No Transcript)
20
Conclusions

• Insiders are informed investors because
• They are contrarians Negative (Positive) CARs
  before buy (sell) trades
• Positive (negative) CARs after buy (sell) trades
• Significant differences between cross-listed and
  domestically-listed companies
• Abnormal returns and
• news impacts
• are significantly smaller for cross-listed firms
• Implication
• bonding contract limits the propensity of
  insiders to trade on insider information
• Coffee (1999), Reese and Weisbach (2002), Doidge
  (2004) and Doidge, Karolyi and Stulz (2004)

21
Questions

• Why do managers still trade before news is
  announced, despite the legal constraints?
• Is the bonding contract not binding?
• King and Segal (2004), Segal (2005) and Licht
  (2003)
• Use other markets
• US domestic vs. UK cross-listed companies
• Other cross-listed companies in US
• Use other news, especially financial analysts
  forecasts
• Market micros-structure effect
• Bid-ask spread adverse selection problem