Globalization and Capital Markets by 2ix90rw8


									Does banks’ corporate control benefit
firms? Evidence from US banks’ control
over firms’ voting rights
Joao A. C. Santos and Kristin E. Wilson

                              Comments by
                             Stijn Claessens
                               World Bank

Findings and General Comments
• Argue that banks’ corporate control lowers agency costs,
  not to force them to borrow at lower costs
• The effect is large: 150 basis points in simple comparison,
  with control variables still 10-50 bps

• New issue for the US (at least); Nice set of data; Many
  robustness tests; and Surprising results!

• Could provide more examples of the mechanisms here at
  work (HP is not a good example)
• While multiple robustness tests, have still some residual
  suspicion that other channels are at play
• Contrast with other countries’ role of banks in commerce:
  control over corporations general bad: why (perhaps)
  different in the U.S.?

Questions on the causes
•       Banks can vote share in trust, have thus control rights without
        cash-flow rights.
•       May have (or gain contingent) control over firms that allows
        them to price loans cheaper
•       Question: is it control, information or value?
•       Control, possibly, but:
    –      Small stakes (0.349% on average, mostly less than 1%)
    –      In US (small) shareholder have very little influence
    –      Doctrine of equitable subordination should make banks reluctant
           to get involved with corporations’ management
•       Information, possibly, but:
    –      These are firms for which much information is available
    –      Chinese walls, officially at least, within banks
    –      Perhaps common analysis, e.g., economies of scale in equity
           investment and loan lending, leading to cost gains?               3
•       Value (due to other factors)
    –     We know subtle “benefits” in financial intermediation can translate
          into large values. And control goes two ways, use and misuse
    –     Banks can be softer to management, be allowed to lend at lower
          rates; even though firms may perform worse, it are the trust
          investors that pay with lower equity rates of return
    –     Could it be that banks get some business value out of running
          trust equity business of firms and then cross-subsidize lending?
•       Examples could help on what is going on
    –     What is it exactly that banks do with their votes?
          •   Do they vote against management more often? They use to (?)
          •   Do they only come in when there is a crucial vote or more often?
          •   Do they vote in situations of near financial distress? And get a better
              share of the value that way? Or
          •   Are banks softer on management (and get a loan out of it)?
•       How do banks compare to other investors that face this?
    –     Do unaffiliated mutual funds, e.g., Fidelity, vote differently?
    –     Is their behavior not the right control for the empirics?
Empirical Framework
•       Empirical setup
    –     Controlling for endogeneity is key in these tests
    –     Paper does a very good job of running all kind of tests
•       Some questions, nevertheless, on empirics
    –     Are spreads all-in-costs? How about re-pricing of loans?
    –     Like to know more on other shareholders and ownership
          structures to see whether bank can exert any influence and
          whether corporate governance is effective in the first place
    –     Can control for the corporate governance practices of firm
    –     Suspect size to play a large role in vote and other factors: more
          interaction effects of size with some variables?
    –     Not sure the choice of vote-non-vote is purely exogenous: could
          have banks choose investments taking investors’ preference into
          account if this is known (which is likely)
    –     What is the shared voting authority? Can one assume the same
          as sole voting authority?

Empirical Extensions
•   The covenants are interesting as they go most directly
    to the agency, moral hazard issues
•   Not clear that they (just) confirm agency issues, could
    still be information story (e.g., learn more about firm by
    holding equity which reduces need for collateral)
•   Can do “system” regressions: explain both spreads and
    covenants (since spreads are function of covenants) as
    functions of control rights
•   Like to know more on the lending structures (e.g.,
    syndicated loans versus single finance) to see
    importance of covenants relative to lending structure
•   Who are the lenders? Are large lenders also large trust
    fund firms? A bias? Do they lend at softer terms as
    these are the better clients? Are they also investment
    banks? Some soft understanding on underwriting?

Policy Issues and Implications
• The paper can expand on implications since one
  always worries when there are control rights
  without cash-flow rights
  – Are equity-holders and banks here both better off or
    does this happen at somebody’s advantage?
  – Is this (another) area where government needs no
    longer to impose restrictions?
• Lessons for other countries
  – Links banking-commerce traditionally viewed
    suspiciously, also in the US
  – When does this work well, if ever, and when not? Are
    conflict of interests taken care of by reputation,
    competition, etc?


To top