Does board independence matter in companies with a controlling shareholder?

Studies have reported valuation discounts for publicly traded companies based in countries that provide weak legal protection minority shareholders.1 Such are often attributed to the ability of controlling shareholders extract “private benefits” come at expense shareholders. Without sufficient deterrents, both incentive and transfer corporate resources themselves personal consumption or gain. These transfers take a number forms, including related-party “tunneling” transactions as well perks and, some cases, outright theft. But under certain circumstances—notably, when their want raise capital by selling shares—the may face stronger reduce this value discount providing credible commitments outside investors forgo diversion resources. Various commitment mechanisms been proposed literature, cross-listing on U.S. exchanges general improvements overall governance systems.2 another possible solution is more effective oversight boards. We recently published study investigated effects appointing independent directors controlled dominant shareholder.3 Using biographical data nearly 8000 799 closely held 22 countries, we found significant positive correlation between fraction board made up directors. Moreover, relation be especially pronounced afford protection—countries where presumably then greatest opportunity increase values submitting greater oversight. Thus, findings our consistent with possibility appointment no ties shareholder can powerful mechanism threat resource from how reliable interpretation, given same power appoint members also power—perhaps if they do too good job—to dismiss them? To address issue, performed several additional tests designed detect monitor actions shareholder. One such test revealed 71% sample sat multiple reasoned appointments likely proxy “reputational capital,” should less willing jeopardize those reputations proving ineffective monitors. As second check whether help resources, frequency was significantly lower larger fractions directors, reduced associated higher values. A third set an shares would most beneficial plan either issue equity behalf firm sell account. The tradeoff faced these circumstances private control benefits. In other words, expected gains share outweigh sacrifice Consistent argument, issued had pages follow, explore each questions raised here depth, report relevant results recent study, present representative case studies. Because appointed dismissed shareholder, question arises perform duties effectively. After all, subject various pressures, dismissal any time, opposing interests believe there three necessary conditions boards monitors First, must able cost diverting benefit. It worth noting not shareholder's perfectly. extent make it difficult costly benefits, preserve Second, Eugene Fama Michael Jensen argued classic paper early 1980s, provided mere existence well-functioning market services directors.4 Thanks market, “human” “reputational” stake lost “when internal breaks down” badly. assess reputational gathered information directorships Finding than 70% over 4000 served boards, concluded robust having enough. Independent well. examining which condition monitoring hold, started finding previous research that, 20 sample, responsibility curb self-dealing legally accountable failing so.5 This mandate gives authority act Along authority, routinely seek obtain assurances will freedom ownership end 2002. predominantly well-developed economies, whose were locate composition. Our initial included 70 largest capitalization country. Before could determine director affiliation, first task identify defined single owner voting rights company least 10% firm's votes.6 Figure 1 provides identity Of all 347 individuals families, 226 privately operating holding companies, 101 financial institutions, 108 governments. last years seen issuance codes, mandates, recommendations, listing requirements worldwide. vast majority them place emphasis independence.7 used publications guide arriving own definition what constitutes director. considered affiliated (i.e., independent) (1) family name (2) employees firm, (3) subsidiary positioned “above” tree, (4) position regardless (5) politicians government agency government, (6) cases foreign employed domiciled country Directors who identified being ways designated “independent.”8 2 unaffiliated of) An inspection chart reveals varies across highest lowest Japan. exceeds proportionate claims cash flows,9 about large disparity cash-flow amplifies incentives exploit shareholders.10 For pyramidal structure determined owned multiplying his along chain tree. example, individual X A, turn 20% B, B 2% (20% times 10%) rights. While customary U.S.-based studies compute Tobin's Q measure value, international datasets generally lack depth needed calculate accurately. primary dependent variable accordingly following variation Q: plus book liabilities assets equity) divided assets. calculated averaged market-to-book ratio 2002 2003. limit effect outliers findings, trimmed top bottom one percent ranked proxy, left final 782 firms. Table 1, firms important characteristics classified countries. extracted only country, differs Japan considerably underrepresented because, among Japanese 16 10 firms, respectively, Italy, contrast, 56, sample. Also investor refer “LEGAL.” quality afforded thought comprising two elements: statutory rules provisions (also called de jure protection) degree enforcement statutes (de facto protection). LEGAL product indices capture protection. “Anti-director rights” index, ranges 0 6, “Law Order” index International Country Risk Guide, rescaled range 10.11 despite focus mostly developed observed ranged low 3.3 Mexico high 50 Canada, U.S., U.K. ownership, composition, (Tobin's Q), averages different terms cross-country variation, percentage (75%) (38%). within-country minimum maximum percentages 45% 93% 0.0% 100% France, Germany, Brazil. first-pass analysis, compared version against index. comparison, into groups eight ≥ 30 group five > 2, nine < 3. Qs varied systematically levels protection, Qs: mean 1.58 meanQ 1.38 3 (p-value = 0.01). analysis confirm accepted view country-level interesting emerged split entire according median monotonically differences means medians statistically (p-values 1.32 1.57 two-thirds. Hellenic Organization Football Prognostics S.A. (OPAP) began life 1958 Greek lottery system operator. Today primarily engaged management, advertisement, operation games. 2003 OPAP ten value. years, continued grow September 2008 become Greece, sales €5 billion, gross profit €1 cap €7 billion. Athens but has 5316 agencies terminals, home abroad, connected online central system. Although incorporated ranks exception companies.12 During past 5 2003–2008 11 members, whom State 75% shares. outstanding (including minority) full equal rights, state commanded flow allowing avoid distortion created structure.13 Starting 2001, listed stock exchange, gradually decrease its firm. By July 2005 45%, May 34%. sense, main impetus behind push level State's decision public. OPAP's disclosure standards, officer remuneration, continue go beyond mandated requirements. unlike code articulates According code, audit compensation committees non-executive chaired member board. December 2002, moreover, comply new law formation department. What evidence standards effective? period examined (2002–2007), related party transactions, red flags shareholder.14 And company's performance impressive. 2003, return 35%, 2004 100%, 50%. outperform (GD) 2006–2008 well.15 boasts remarkable profitability ratios, ROAs 54% 2007 51% ROEs 105% 98% 2008. add rigor descriptive statistics, estimated series regressions show Q.16 key coefficients regression reproduced 2. Column presents “base case” regression. serves compare However, excludes interest, (“INDDIR%”). prior studies, nationwide coefficient (and p-value 0.01), suggesting increases There indication enhances (though corresponding only). subsequent regressions, base remained essentially unchanged. All shown positively correlated shareholder.17 Having base-case regression, added INDDIR%—again, directors—to column table, INDDIR% interpreted saying (in members) 4% ratio. thus, $1 billion value), translates $40 million Next, limits “linearity” relationship, replaced INDIRR% natural log nonlinear specification INDDIR%. (reported 3) 0.02), adjusted R2 increased relative linear specification. better “fit” logarithmic although decreasing rate thus adding fifth adds fourth). sum, suggest intent raising consider directors.18 next turned attention issue. ln (INDDIR%) strong valuable offers because potential increasing greater. plausible counter argument—that little absence protective environment, fundamentally mercy Mobistar entered mobile phone operator Belgium joint venture Telinfo France Telecom 1996. Telecom's wireless Orange. French making ultimate Mobistar. Up until Orange maintained slightly above 50% owns civil limited self-dealings shareholder.19 Furthermore, Belgian Code Corporate Governance did December, 2004, is, year after collected. company, however, committed creating best environment management succeed. Both chairman CEO local, dominant-shareholder. whole displayed considerable independence uncharacteristic companies. spite Telecom, 62% unrelated stake. (2002–2003), (unlike 17 sample). With through 2007, officers existence, expanded aggressively. Its 3% 33% 1996 became Bel20 accounts 80% $274 market. GPRS Belgium. At ratios average industry.20 achieve success initially debt markets. relied heavily finance expansion 3G network 2001–2002. (In resorted secondary offerings year.) Maximizing newly considerations practice high-quality governance.21 needs external funds decreased since February repurchasing remarkably falling, appreciated 20%, gained further 40% 2004. Since then, price relatively stable outperformed 25% Mobistar's ratio, assets, 2.63 3.85 3.0 then. hovered around 1.5. arguments, re-estimated subsets companies: (those 20) 30). columns 4 4, implies, 90% like (with 3.3), double (from 0.68 1.29). degrees reliably zero 0.55). support reducing Given assumption equity, out examine seasoned (SEOs) during 2002–2004, 198, four, undertook SEOs $10 million. pass, not, INDDIR%, groups. 56.1%, while non-issuing 50.2% (The difference Next intended investigate systematic model 1/0 indicator SEO (not offerings).22 3, (at 0.01 less),23 supporting conjecture determinant composition alleged advantage position. frequently referred tunneling.24 cannot directly observed,25 assumed much activity accomplished (RPTs). Strong wealth RPTs preventing clearly holders. Following line reasoning, hypothesized occur directors; occur, advantageous Though way knowing RPTs, observe requires periodic filings. classify transaction RPT, borrowed classification scheme adopted studies.26 types dealings RPTs: acquisition and/or shareholder; asset swaps firm-affiliated loan relief purchases merchandise shareholder.27 search filings yielded conducted 148 49.4%, 53.2% 3.8%, 0.14; 7.1% 0.03). factors affect likelihood logit occurrence RPT (INDDIR%). negative significant, implying reduces RPTs. typical everything else fixed 60% odds observing 19%, approximately 30%. Financial Times, “Keyence people; engineers, great deal.”28 Keyence Corporation direct employs 1500 personnel (most engineers) deliver custom automation solutions factories 32 founded 1974 Takemitsu Takizaki. Keyence's delivering innovative engineering products. grown annually, average, reaching $2 mark. 2000, appeared Nihon Keizai Shimbun's yearly ranking “Top Ten Most Excellent Companies Japan.” culture,29 played role. world keiretsu predominant form organization largely relations, 2003) seven fewest sample; fully founder. one-third seats occupied shareholder—the smallest participate transactions. long-established paying part dividends, works diversion.30 fast growth forced resort issues. announced sizeable offerings. November $111 shares, followed $347 June Only time frame. issues Mr. Takizaki's 25%. motive maximize proceeds 4.5 1.96 Further, October moved erratically saw decline still Tokyo Stock Exchange Index 12%. BusinessWeek's “1000 Best Valued Companies.” supplement (INDIRR%) random 2002–2003 RPT. variables except excluded (board size). presented 0.06), 0.04). Thus appear even environment. analyses interpretation limiting disadvantageous indicate so shareholders, But, suggest, 90%, expect see almost 15% 1.41 1.60). Nevertheless, enough loss suggests say U.K., 1.90 composed moderate protection? Can benefit board? India, 20, 1.80. equivalent country's score 40 (the Australia). effective, though probably complete, substitute strengthening big difference, particularly weak.