The%20Dark%20Role%20of%20Investment%20Banks%20in%20the%20Market%20for%20Corporate%20Control

1
The Dark Role of Investment Banks in the Market
for Corporate Control

• A. Bodnaruk, M. Massa A. Simonov

2
Why MA arb and inv banks?

• Market for corporate control should discipline
  management. But it does not
• Long-term performance of merged companies are
  mixed at best. Usual explanations are hubris and
  empire building
• In 1990-es we had a lot of MA activity that
  resulted in massive distraction of shareholder
  wealth.

3
Investment banks in MAs

• Advise to the bidder on the strategy
• Evaluate the assets of the target
• Help to execute transactions
• Certify the deals
• Provide assistance in the negotiations
• As many as 17 investment banks were working on
  Arcelor/Mittal deal (FT,June 28th 2006)
• Servaes and Zenner (1996) Hefty commissions and
  fees (0.5 -1 of target value) are justified by
  lower transaction costs and superior expertise

4
A small detail

• Investment banks are privy to the information
  about the upcoming bid
• Do they exploit it?
• If so, is it a service to a target (toehold) or
  greed?
• If it is greed how far does it go?
• Just take positions and hope for the deal to go
  through or
• Affect the outcomes of the deals (to maximize
  their profits)?

5
A little bit of mechanics of MA Arb

• Upcoming bids are difficult to predict
• Arbitrageurs take positions 1 or 2 days after the
  bid announcement
• Miss most of the price increase

6
MA Arb Primer

• Profitsp(success) spread (1-p)LNN

7
High SR, but not mkt neutral
8
The rise of MA arb in Inv. Bank

• First time documented Kidder, Peabody Co in
  mid-80es set up MA Arb (run by Martin Siegel)
  that provided lion share of overall profit for
  the whole KP.
• Now most banks do have proprietary trading desks
  that are involved in MA arb.

9
Good old times?

• Book published in 85 I considered whether to
  talk about behind the scenes maneuvering and
  smoke-filled rooms, but I decided I wanted to do
  a serious book on arbitrage
• Who wrote it?

10
Is it conflict-free now?

• In 2003 SEC fines Deutsche Bank 750,000 for
  hiding a conflict of interest when it voted (in
  Compaq/HP merger). A judge who heard evidence
  during a hearing in 2002 said "This fact raises
  clear questions about the integrity of the
  internal ethical wall that purportedly separates
  Deutsche Bank's asset management division from
  its commercial division."
• NY Times, Aug. 26, 2006
• it is undeniable that brokerage firms, with
  their varied businesses all under one roof,
  remain particularly well-positioned to capitalize
  on inside information In a July 7 speech, Hector
  Sants, managing director of wholesale and
  institutional markets at the F.S.A., described
  why his focus was shifting to institutions. Our
  spotlight will shine in particular on
  relationships between investment banks and their
  clients, he said, because we believe the risk
  of market abuse is highest where a client can be
  made an insider on a forthcoming deal.

11
What we show

• Investment Banks advising the bidder are taking
  positions in target.
• In suspicious deals premiums are larger
• Suspicious deals have higher probability of
  being completed they also are more likely to
  have target termination fees.
• Positions that IBs take are very profitable and
  cannot be replicated based on publicly available
  info
• We also show some evidences that suspicious
  deals are bad in the longer run
• Overall our evidence are consistent with not just
  benefiting, but engineering deals to their own
  advantage

12
Literature MA arb.

• Larcker and Lys (1987) superior ability to
  predict offer outcomes leads to their abnormal
  returns.
• Cornelli and Li (2002) and Gomes (2001) suggests
  an active role for arbitrageurs. The information
  advantage that an arbitrageur possesses arises
  from her own holdings and that these holdings
  influence offer outcomes and deal
  characteristics.
• Mitchell Pulvino (2001), Baker and Savasoglou
  (2002) a is there, but they are not mkt neutral.
• Rau (2000) Contingent fees are correlated with
  past market share (and successful completion) and
  correllated with future performance
• Matvos Ostrovksy (2006) many institutions have
  positions in both target and bidder and benefit
  from MA arb.

13
Literature Conflict of interest

• Acharya and Johnson (2005) lending banks use
  private information regarding corporate clients
  to trade credit default swaps.
• Irvine, Lipson and Puckett (2004) institutional
  investors receive tips regarding the content of
  forthcoming analysts reports
• Ritter and Zhang (2006) lead underwriters
  allocate hot initial public offerings to
  affiliated funds.
• Ellis, Michaely and OHara (2000) NASDAQ market
  makers belonging to a financial group support the
  stock price of those firms whose IPO has been
  underwritten by the investment bank belonging to
  the group.

14
Good or Bad?

• IBs can reduce cost of acquisition by using their
  expertise (Servaes Zenner 96).
• It is just so happen that IB and Asset Management
  arms pick up the same companies (the same
  analysis leads to the same conclusion).
• IBs can take position in the target on behalf of
  the bidder (toehold).

15
Good or Bad?

• Information Hypothesis (or informed trading one)
  Asset Management Arm uses MA private info to
  take positions
• Conditioning Hypothesis IB plays active role in
  selecting the target and making bidder to agree
  to pay higher price.

16
Example Compaq/HP/DB timeline

• March 15, 2002. Officials of DB asset management
  division, which owns 17 mln HP shares, vote
  against the deal.
• March 17. In a voice mail Fiorina tells CFO Bob
  Wayman they need to do "something extraordinary"
  to win over DB.
• Mar. 17-19(?). HP executives confidentially hire
  DB investment bankers to provide "market
  intelligence" during the proxy fight, agreeing to
  pay them 1 mln, with a 1 mln bonus if the
  merger goes through. Then they asked DB IBs to
  intervene. The bankers contact Dean Barr, a top
  exec from DB Asset Management. On March 19 vote
  was changed.
• At the same time, DB accumulated 5.268 mln shares
  in COMPAQ. Rough profit estimate20mln (fine was
  0.75mln)

17
Data

• CRSPCOMPUSTAT
• SDC newspaper clipping from FACTIVA
• Spectrum IH and 13F
• Link between 13F and SDC was created
• Info used SEC (Adviserinfo), Morningstar, web.

18
BRAND
19
Brands are changing
Bought by MK 1989-94
20
Insiders

• A brand is labeled as non-insider arbitrageur if
  its holdings in targets go from zero to positive
  in at least 20 deals in our sample following bid
  announcement
• Insider arbitrageurs include brands which either
  advise to acquirer (insider to acquirer) or to
  target or provided a loan to a target no more
  than three years prior to first bid (insider to
  target).

21
Are IBs smart or get help?
22
Can it be that they are acquiring toehold?

• Unlikely. Conditional on taking position, it
  should be large (Eckbo Betton 2001 8.5,
  Jenter 2006 17). Our position is about 0.6...

23
Potential insiders variables

• Ownership by Pot. Insiders is the fraction of the
  company which is held by brands acting as
  advisors on at least 5 deals in our sample.
• Ownership by Pot. Insiders is the log of one
  plus the Dollar value of the stake of the
  potential insiders holding positions in the
  company. Prices are measured at t-62
• Potential insiders share of arbitrage is the
  Dollar value of the stake of the potential
  insiders holding positions in the company,
  divided by the aggregate arbitrage capital in the
  market.
• Ln(N Potential Insiders) and Ln(N Deals) measures
  are the logarithm of the number of potential
  insiders in the company and the logarithm of the
  combined number of deals these brands advised
  over the sample period

24
Actual insiders

• Ownership by Advisor the fraction of the
  potential target equity held by the advisor to
  bidder
• Ownership by Advisor - i.e., the log of dollar
  value (plus one) of the advisory banks
  position.
• Ownership by Advisor Dummy

25
Other variables

• Total changes in arbitrage capital (mkt-wide)
  (Cornelli and Li, 2002, Baker and Savosoglu,
  2002, Hsieh, 2001)
• Institutional ownership (Stulz, Walkling and
  Song, 1990)
• ROE, B/M, Size, Sales Growth, Accounting
  Liqudity, P/E, D/E for both target and bidder
• Industry herfindahl, momentum, volatility
• Contractual features (as in Officer)

26
Probability of Becoming Target

• Increasing Ownership by potential insiders by one
  standard deviation increases probability of being
  takeover target from 1.8 to 2.1, or by 17 pct
  pts.
• Increasing Ownership by potential insiders by
  one standard deviation increases probability of
  being takeover target from 1.8 to 3.5, or by 92
  pct pts.
• Increasing Potential insiders share of arb
  capital by one standard deviation increases
  probability of being takeover target from 1.8 to
  2.5, or by 40 pct pts.

27
Matched sample
Probability of becoming a target increases from
4.2 to 6.1
28
Premium

• We follow Schwerts (2000) definition of Target
  Abnormal Return Premium. We use both one-factor
  and four-factor adjusted premiums.
• Loadings are determined using 253 trading days
  ending at day -64 (i.e., trading days (-316,
  -64)).

29
Premium

• An increase of one standard deviation in the
  advisor stake (corresponding to approximately a
  0.53 ownership of the firms) increases the
  target firms premium from 25.5 to 27, or by 6
  percentage points.
• It goes to 28.4 and 31.1 in the case of
  Ownership by Advisor to Acquirer and Ownership
  by Advisor to Acquirer Dummy respectively.

30
What about trading strategy?

• Let us look at actual profitability of the
  positions IBs are taken. We compare
• Return on positions held by actual advisors
• Return on positions taken 1 day after deal
  announcement (separately for deals that are not
  held and held by advisors)
• Returns on Cash/Stock deals

31
Descriptive stat
32
Fama-French-Carhart regressions

• a is about 4-4.5 per month
• It seems that there is no systematic risk in
  advisor takes positions

33
Advisors always make money
34
and a lot of money
35
Strategy based on prob of becoming target

• We can build the strategy looking at medium size
  companies (deciles 4-9) using probit of becoming
  target with actual and potential insiders and
  selecting top X stocks (by predicted prob)

36
Four-factor model
37
Next step

• Are IBs just making money by taking positions or
  do they go further? Would they engineer the deal?
• We will show that they make completion of the
  deal more likely, they also use contractual
  features that increase success
• Usual disclaimer about potential reverse
  causality applies
• We present weak evidence that the deals are
  suboptimal for the companies

38
Target Termination Fee

• Target termination fee (fee payable by target to
  bidder if target walks away from mutually agreed
  deal) enhances chances of deal being complete
  (Officer, 2002)

39
Probit estimates

• Even controlling for all other contractual
  features results hold
• An increase of advisory stake by one standard
  deviation increases probability to have target
  termination fee from 40 to 44.7, or by 12 pct
  points.

40
Success

• Deal non-completion is the biggest risk MA
  arbitrageur is facing.
• In deals with holdings by advisors probability of
  completion is higher by almost 6!

41
Probit

• The results survive control variables
• An increase in the advisory stake of one standard
  deviation increases the probability of success
  from 77.8 to 80.1.
• It goes to 81.6 and 83.4 in the case of
  Ownership by Advisor to Acquirer and Ownership
  by Advisor to Acquirer Dummy, respectively.
• So, 10-25 of RISK of non-completion are gone!!!
• Alternative instrument target term fee with
  Ownership by Advisor variables. Actually, works
  similarly.

42
Profit Margins next FY after completion

• Profit Margins are lower by about 2
• Similar results holds for ROA, ROE

X
X
43
Whos the biggest loser?

• Which bidders are most likely to be taken
  advantage of?
• We construct measures of bidder sophistication
  related to the prior experience in financial
  markets and the relationship with advisor
• these include of bond equity issues, of
  MAs conducted in the previous 3 years, both
  unconditionally and with the help of current bid
  advisor
• Result Companies with low experience and no
  prior relationship with deal advisor are
  suffering the most.

44
Other results

• Similar to term. fees result was obtained for
  collar (but smaller econ. magnitude).
• In the deals with large distraction of
  shareholder wealth, more than half has IB
  position in the deal.

45
Conclusion

• We provide evidence that advisors to the bidders
  have positions in the target before the deal. The
  existence of a direct stake of the advisor to the
  bidder increases the probability that the deal is
  successful as well as the target premium. We
  explain these findings in terms of insider
  trading of the advisory bank.
• Our findings suggest that advisors not only take
  advantage of their privileged position by getting
  involved in MA Arb, but also by directly
  affecting the outcome of the deal in order to
  fetch a higher capital gain from their positions.
  
• These results provide important insights on the
  conflicts of interest that affect financial
  intermediaries that can both advise and invest in
  the equity market.

46
It is good to be the king investment banker!