Discussion Real Effects of Different Types of Ownership

Transcription

1 Discussion Real Effects of Different Types of Ownership Shareholders at the Gate? Cross-Country Evidence on the Role of Institutional Investors in M&As (by Matos, Ferreira and Massa) Pension Reform, Ownership Structure and Corporate Governance: Evidence from Sweden (by Giannetti and Laeven) Corporate Governance, Regulation, and Bank Risk Taking (by Laeven and Levine) by José Luis Peydró European Central Bank This discussion reflects the views of the discussant only and not those of the ECB or Eurosystem

2 Outline of the discussion Common discussion of the 3 papers Discussion in more detail of each paper

3 This conference is on Asset Management, Private Equity Firms and International Capital Flows: Their Role for Financial Integration and Efficiency This session is on Real Effects of Different Types of Ownership 1st Change in control and value Ownership 2nd Value 3rd Risk Integration Efficiency By type: mutual funds, pensions funds, banks, non-financial firms By nationality: nationals vs. foreigners (different classes of asset management and international capital flows)

4 Normal paper in the conference: Private equity LBOs Change in control Value Debt ( Risk) This session: General ownership (institutional investors, ) Much bigger than PE, HF Not limited to one operation (LBOs, voting ) General impact on value, risk

5 This session is more general But the econometric identification is more difficult Ownership efficiency and integration Problems for causal inference: Reverse causality Endogenous choice Omitted variables Solutions: Panel Interaction variables IV Natural experiments, exogenous variation in ownership

6 Results Matos, Ferreira, and Massa: IO driving force in market for control (M&As) IO Volume M&As & Cross-border M&As Giannetti, and Laeven: Pension reform: Nice for identification Pension funds Firm Value depending on PF To increase value, PF large and independent Laeven, and Levine: Large owners Bank Risk Taking Depends on LO in board and regulations

7 Policy implications Matos, Ferreira, and Massa: Key result for financial integration and efficiency If IO can play, they can bring financial integration (Cross-border M&As) and they can create value Giannetti, and Laeven: Pension reform: very important changing corporate governance and incresing firm value and performance Laeven, and Levine: Key result for banking regulation and supervision Capital requirements don t matter If investor protection is strong, LO don t matter

8 Matos, Ferreira, and Massa About the panel: Why not interactions? Effect of IO on M&As should depend on investor protection, regulations (strategic substitutes or complements)? (causal inference and policy implications) IV: no first stage regression (weak instruments), the theoretical validity of I, all instruments together in one regression?

9 Matos, Ferreira, and Massa Very interesting result that common shareholders in T & A gain a lot with the deal If so, and based on your tables, you could try: P(target)= F(IO[+], IO(Common)[+]) where the counterfactual could be all firms in same industry and common is conditional on acquirer P(acquirer)= F(IO[-], IO(Common)[+]) and this relation should depend on regulation IO(Common)&premium target bigger loss in acquirer

10 Giannetti, and Laeven The part closer to a quasi-natural experiment is the first part of the pension reform, the second part is not so much Interpretation of the main result: Price pressure The reduction of AP4 is after the crash of the bubble in March It implied a crisis and maybe bigger firms the ones that are maybe more international suffered more (industry fixed effects not perfect) If public pension funds sell, the potential implicit garantee (a put) of having public shareholders dissapears, hence a reduction in value

11 Giannetti, and Laeven Since you have all the owners, the main result for the AP4 should be stronger or weaker depending who bought the shares, you can test this Interactions effects? The effect of PF on value should depend on the need for monitoring (wedge, CF rights bigger shareholder ) PF dependent on banks should not always be bad; they should be bad if and only if they belong to the large shareholders of the firm?

12 Laeven, and Levine Econometric identification 1 point: level of ownership bank risk (changes in ownership would be much better ) Level of ownership and management structure should be predetermined, they are based on 2001 and risk (the LHS variable) is from The only strong instrument is CF(others), but if the hypothesis you test is true, then CF(others) cannot be an instrument since CF(others) risk (others) risk Measure of risk is based on accounting, why not EDF from KMW?

13 Laeven, and Levine Large owner bank risk taking. Even with very large owners (CF>20%) you find that risk taking is higher, but maybe value is lower, i.e. entrenchment vs. risk-taking? Does z-score reflects risk taking or stability (value and risk)? The results with standard deviation are less significant In countries with strong investor protection, large owners do not affect bank risk. Why? Small countries are overrepresented (10 US, 10 Danish and 10 South African banks) and there are fewer banks with strong investor protection Large owners is very extreme, maybe not very important in countries with strong investor protection, other more complex ownership structures may influence bank risk taking