Multiple Blockholders, Price Informativeness, and Firm Value

Transcription

1 Multiple Blockholders, Price Informativeness, and Firm Value Alex Edmans, Wharton Gustavo Manso, MIT Sloan Notre Dame Finance Seminar October 12, 2007 Edmans and Manso Multiple Blockholders October 12,

2 Introduction What is the optimal blockholder structure that maximizes the e ectiveness of corporate governance? Traditional view: blockholders govern through intervention ( voice ) Few, concentrated stakes are optimal to minimize free-rider problems and maximize intervention incentives But most rms have multiple small blockholders (Zwiebel (1995), Barca and Becht (2001), Faccio and Lang (2002), Maury and Pajuste (2005), Holderness (2007)) This paper provides a potential justi cation for multiple blockholder structures Edmans and Manso Multiple Blockholders October 12,

3 Introduction (cont d) MBs increase the e ectiveness of a second governance mechanism: trading ( exit ) Managerial e ort improves fundamental value, but manager is paid according to stock price Trading causes prices to re ect fundamental value, rewarding e ort ex post Dynamic consistency issues: once e ort has been exerted, blockholders are only concerned with maximizing trading pro ts A single blockholder will strategically limit her order, reducing price informativeness Multiple blockholders trade competitively, as in a Cournot oligopoly * Here, co-ordination problems help, by serving as a commitment device to reward the manager ex post Edmans and Manso Multiple Blockholders October 12,

4 Related Literature Zwiebel (1995), Bennedsen and Wolfenzon (2000): control contest motivations for MBs Di erences with Edmans (2007): Considers both voice and exit, and the trade-o s between them Fundamental agency problem is shirking, not myopia Derives multiple blockholders as an optimal structure, and analyzes the determinants of the e cient number Edmans and Manso Multiple Blockholders October 12,

5 The Model Manager owns α, I blockholders collectively own β (i.e. β/i each) t = 1: manager exerts e ort a, each blockholder exerts e ort b i (all at unit cost) Firm value: ev = φ a log a + φ b log i b i + eη a is private (critical), b i is public (non-critical) t = 2: trading by blockholders and liquidity investors Each blockholder observes ev perfectly and demands x i (ev) Noise traders demand eε N(0, σ 2 ε ) Market maker observes order ow ey = i ex i + eε and sets ep = E [evjey] Manager maximizes αep a Blockholder maximizes individual trading pro t + value of shares - cost of e ort Edmans and Manso Multiple Blockholders October 12,

6 The Trading Stage Proceed by backward induction: take a and b i as given Unique linear equilibrium is symmetric: x i (ev) = γ(ev φ a log a φ b log i b i ) 8i p(ey) = φ a log a + φ b log i b i + λey, Each blockholder s trading pro ts are 1 p I (I +1) σ η σ ε Price informativeness is negatively related to σ 1 Var(evj p) = 1 I +1 σ η Edmans and Manso Multiple Blockholders October 12,

7 The Action Stage Optimal actions: I a = φ a α I b i = φ b β I 1 i b i = φ b β. I i b i is decreasing in I Blockholders exert positive externalities on the rm, which they do not consider Too little intervention: standard free-rider problem Edmans and Manso Multiple Blockholders October 12,

8 The Action Stage (cont d) a is increasing in I Blockholders exert negative externalities on each other, which they do not consider Too much trading * Reduces informed trading pro ts, hurting blockholders in aggregate * Increases price informativeness But rm value does not depend on informed trading pro ts, but instead price informativeness * Co-ordination problems commit to too much trading, creating a dynamically consistent reward mechanism for managerial e ort Edmans and Manso Multiple Blockholders October 12,

9 The Optimal Number of Blockholders h i Optimum for rm value: I = max 1, φ a φ b φ b I depends on trade-o between e ect on blockholder e ort ( voice ) and managerial e ort (via exit ) φ b depends on nature of blockholders expertise * High if forward-looking ( prospective ) information, e.g. venture capital * Low if backward-looking ( retrospective ) information, e.g. rentier capital * High if strong control rights (holding β constant), e.g. foreign blockholders φ a depends on manager s scope to improve rm value * High in growth industries, low in regulated industries Weak governance increases both φ a and φ b and thus has ambiguous e ects Edmans and Manso Multiple Blockholders October 12,

10 The Optimal Number of Blockholders (cont d) Social optimum considers: Cost of managerial e ort (reducing I soc relative to I ) Cost of blockholder e ort (increasing I soc relative to I ) Private optimum considers: β% of increase in rm value Cost of blockholder e ort (increasing I priv relative to I ) Informed trading pro ts (reducing I priv relative to I ) Edmans and Manso Multiple Blockholders October 12,

11 In Progress... E ect of information asymmetry Likely to reduce potency of exit and thus I E ect of liquidity Currently, σ ε is irrelevant, but may become signi cant if costly information acquisition Likely to increase potency of exit and thus I Asymmetric blockholders Further suggestions? Edmans and Manso Multiple Blockholders October 12,