Asymmetric Info, Signaling, Pecking Order

1
Asymmetric Info, Signaling,Pecking Order
2
Overview

• In a simple world everyone is equally informed
• In the real world some parties (ie insiders) may
  have more info than outsiders
• Rational investors will anticipate insiders
  having more information
• This may affect optimal capital structure and
  dividend policy

3
Quick Question

• Green technologies are a hot topic
• I am the CEO of a firm producing ethanol fuel. I
  am aware that ethanol is too expensive and
  unlikely to catch on, do I raise money through
  debt or equity?
• I am the CEO of a firm producing solar energy. We
  are very close to a breakthrough doubling energy
  output, do I raise money through debt or equity?
• I am a rational investor. I observe a hydropower
  startup raising money through debt, what should I
  infer about their technology?

4
Signalingwith Dividends

• Proposed by Bhattacharya (1979)
• Risk neutral world (5 discount rate)
• Otherwise return on equity depends on leverage
• Could use state prices or WACC
• You (the entrepreneur) need to raise 10 for a
  project
• The market believes the project will pay 5 w/ p,
  20 w/ 1-p (assume p.5)
• You believe the true probabilities are q and 1-q
  (q may be different from p)

5
Market Valueof Debt

• Present value of debt with face F is
• Bpmin(F,5)(1-p)min(F,20)/1.05
• If Flt5 BF/1.05
• If 5ltFlt20 Bp5(1-p)F/1.05
• If Fgt20 (we dont care about this case)
• Bp5(1-p)20/1.05
  
• Total value of Equity
• Epmax(5-F,0)(1-p)max(20-F,0)/1.05

6
Market Valueof Equity

• You must raise remainder with equity. That
  remainder is 10-B
• What share a of the firms equity must be offered
  to the market? Must satisfy
• 10-B aE apmax(5-F,0)(1-p)max(20-F,0)/1.0
  5
• If Flt5
• If 5ltFlt20
• If Fgt20 equity is worthless (we dont care about
  this case)

7
(No Transcript)
8
(No Transcript)
9
QuickQuestion

• Does MM hold so far? Can you confirm or refute
  this from the graph?
• The share of equity that needs to go to outside
  equity investors decreases with the face value of
  debt. Why is this?
• Why does a0 at F16? Can you predict (without
  doing any calculations) what the market value of
  debt is when F16

10
Value to Entrepreneur

• The entrepreneur receives the share of equity not
  promised to equity investors, that is 1-a
• If Flt5 entrepreneurs share
• If 5ltFlt20 entrepreneurs share
• Note that the expression for 1-a above depends on
  the markets belief p
• This is because a is set by the market and
  outside investors agree to invest based on what
  they believe about the firm
• However the entrepreneurs belief about the total
  value of equity depends on q because
• Her total payout is (1-a)E where the first
  piece depends on p but the second piece on q

11
Value to Entrepreneur

• Let E be the value the entrepreneur believes
  equity is worth
• Eqmax(5-F,0)(1-q)max(20-F,0)/1.05
• If Flt5 entrepreneurs perceived payoff is
• If 5ltFlt20 entrepreneurs perceived payoff is

12
(No Transcript)
13
NPV of Outside Equity (if insider is right)

• Suppose insider (or old equity or entrepreneur)
  is right ? q is the true probability even though
  market uses p
• The NPV of new equity is aE- (10-B)
• The 1st piece is what they get, the 2nd piece is
  what they paid in
• a and E calculated just as before
• If Flt5 new shareholders payoff is
• If 5ltFlt20 new shareholders payoff is

14
(No Transcript)
15
Intuition

• Suppose market is unconditionally correct. If you
  look across all firms in the economy, p of them
  will produce 5M and (1-p) will produce 20M
• Suppose insiders have more information than
  outsiders. That is the insider of each firm knows
  that this firms probability of producing 5M is q
  which is not always equal to p
• It is said the insiders signal is more precise
• Of course, by our first assumption, the average q
  across all firms must equal to p
• If insiders do not let their knowledge of q
  determine how their raise equity, than outsiders
  will lose on some deals (qgtp) and win on other
  deals (qltp), but will break even on average
• Note that outsiders NPV is linear in p-q so on
  average this is zero
• However, do you think insiders will not let their
  knowledge influence their choice of capital
  structure?

16
Sly Insiders

• Suppose ½ of all firms have q.1, others q.9
• When q.1 (good firm), insider is better off
  raising as much safe debt as possible so that she
  can take advantage of the cheap equity
• In this simple example, risky debt affects
  insider in exactly the same way equity does, in
  general risky debt looks like a mix of safe debt
  and equity
• When q.9 (bad firm), insider is better off
  raising as much equity as possible (that is
  raising no debt at all), so that she can unload
  expensive equity on outsiders

17
Astute Outsiders

• But outsiders are not idiots!!!
• At least not in the model, we will see evidence
  from real world soon
• Outsiders dont know type of firm, but they know
  that they dont know and that insiders do know
• If an outsider observes firm raising all equity,
  she must assume its a bad firm and deduces q.9 ?
  Values firm accordingly
• If an outsider observes firm raising all debt,
  she must assume its a good firm and deduces q.1
  ? Values firm accordingly

18
Full Info

• Bad firm raises all equity
• This is impossible since alt1
• This firm will not receive funding
• This is because this firm has NPVlt0
• Good firm raises safe debt F5, rest with equity

19
Crafty Insiders

• But the managers of the bad firm are not
  idiots!!!
• They will pretend to be a good firm by raising
  safe debt since they cannot get funding otherwise
• But the outsiders are not idiots!!!
• They will no longer offer a41 but account for
  some bad firms and require a larger share

20
PoolingEquilibrium

• All firms raise safe debt F5, rest with equity.
  Since outsiders do not know which is which, they
  use unconditional probabilities
• Good firms are hurt by this and bad firms benefit
• Suppose the good state paid 15 instead of 20.
  Then there would be a market failure a110 and
  no firms would get funded!
• What can be done about this?

21
Signaling

• What if good firm could do something that was so
  outrageous the bad firm refused to imitate it
• For example, what if the good firm decided to
  burn money?
• Lets say firm has C/1.05 cash today (so that it
  is sure to be worth C tomorrow)
• Below we will go through same numerical example
  as above but firm has cash

22
(No Transcript)
23
Market Valueof Debt

• Present value of debt with face F is
• Bpmin(F,5C)(1-p)min(F,20C)
• If Flt5C BF/1.05
• If 5CltFlt20C Bp(5C)(1-p)F/1.05
• If Fgt20C (we dont care about this case)
• Bp(5C)(1-p)(20C)/1.05
• Total value of Equity
• Epmax(5C-F,0)(1-p)max(20C-F,0)/1.05

24
Market Valueof Equity

• You must raise remainder with equity. That
  remainder is 10-B
• What share a of the firms equity must be offered
  to the market? Must satisfy
• 10-B aE apmax(5C-F,0)(1-p)max(20C-F,0)
  /1.05
• If Flt5C
• If 5CltFlt20C
• If Fgt20 equity is worthless (we dont care about
  this case)

25
Value to Entrepreneur

• The entrepreneur receives the share of equity not
  promised to equity investors, that is 1-a
• If Flt5C entrepreneurs share
• If 5CltFlt20C entrepreneurs share
• Note that the expression for 1-a above depends on
  the markets belief p
• This is because a is set by the market and
  outside investors agree to invest based on what
  they believe about the firm
• However the entrepreneurs belief about the total
  value of equity depends on q because
• Her total payout is (1-a)E where the first
  piece depends on p but the second piece on q

26
Value to Entrepreneur

• Let E be the value the entrepreneur believes
  equity is worth
• Eqmax(5C-F,0)(1-q)max(20C-F,0)/1.05
• If Flt5 entrepreneurs perceived payoff is
• If 5ltFlt20 entrepreneurs perceived payoff is

27
PoolingEquilibrium

• It is still the case that good firm better off
  issuing safe debt (face 5C), bad firm issuing
  equity
• It is still the case that there are no idiots.
  Bad firm imitates good firm, everyone issues debt
• All firms raise safe debt F5C, rest with
  equity. Since outsiders do not know which is
  which, they use unconditional probabilities
• Check that NPV of old equity is equal to NPV of
  old equity from old example NPV(C)

28
BurningMoney!

• Manager of good firm (q.1) says
• I will go to Trafalgar Sq and burn my cash
• Bad firm (q.9) will not dare do such a thing
• Investors will realize this and set a41 just as
  in simple world where q is known and C0
• Suppose C2M. Can this work?

29
SeparatingEquilibrium

• If good firm does nothing, old equity gets
• If good firm burns money and investors believe
  it, old equity gets (.4141full information a)
• Good firm would like to do this!
• If bad firm tries to imitate good firm, old
  equity gets
• If bad firm does nothing, old equity gets
  C2/1.051.9
• Bad firm does not want to imitate good firm
• Good firms signal is credible
• Signaling works!!!

30
Intuition

• A bird in the hand is worth two in the bush
• For bad firm, this is true, it is very unlikely
  to catch even one bird in the bush
• For good firm, this is not true, it knows that
  even if it lets the bird in the hand go, it is
  likely to get both birds in the bush
• Investors realize that the only firm that is
  willing to throw out a bird in the hand is one
  that is good at catching them in the bush

31
Realism?

• Burning money in Trafalgar Sq unrealistic?
• Think back to Taxes lecture. What do you think
  companies are doing when they choose to payout
  with dividends rather than share repurchases?
• Lets redo above example and assume we are only
  burning TC amount of money, and keeping (1-T)C.
  Signaling will work if
• Good firm finds it optimal to burn
• Bad firm finds it optimal not to burn
• Thus with a dividend tax rate of 44, firms could
  use dividend payments as signals
• I could easily redesign example with more
  realistic tax rates. If C4 than 22ltTlt30 does
  the trick
• Remember, difference between div. and cap gains
  is about 15

32
CostlySignaling

• Costly signaling is a way to distinguish yourself
  from your competitors
• Quick Question
• Suppose this class adds nothing to your future on
  the job skills
• However, you are smarter than the person to your
  left and are more likely to get a high score
• Employers cannot tell you apart from your
  neighbor, but can see your score
• Look around. Based on these assumptions, raise
  your hand if it is worth it for you to be here?
• How do you feel about being graded on a curve?
• 2001 Nobel Prize Signaling with Education
  Spence (1973)

33
Intuition

• Here good firms used costly dividend payments to
  signal
• This may explain why firms choose to use
  dividends for payouts when share repurchases are
  cheaper
• Remember that debt has a tax advantage that is
  too strong to be explained by trade-off theory
• This means carrying too little debt is costly
• Could design similar example where firm carries
  too little debt as a signal about its quality
• Possible explanation for capital structure

34
Equilibrium

• Pooling Equilibrium All agents in model act in
  the same way, cannot be distinguished, receive
  the same contract
• Separating Equilibrium Agents can be
  distinguished or separated into types based on
  their actions or characteristics
• It must be the case that one type of agent does
  not find it optimal to act like another type

35
Signalingwith Debt

• Again, two types of firms q.1 and q.9 (½ of
  each kind). Need investment 10M to get payoff 5M
  w/ q and 20M w/ 1-q
• Again, firms have C/1.05 cash w/ C3M
• This time there is a corporate tax rate T20 on
  equity
• For simplicity, assume risky debt is not an
  option (ie bankruptcy costs astronomically high)
  so firm can raise at most 5M in debt

36
Full Info

• Good firm raises C5 face value debt
• a23 for good, 208 for bad firm
• bad firm again will not be funded since NPVlt0
• Value to old equity
• Good firm raises only equity
• Value to old equity
• Under full info, good firm prefers debt because
  of tax savings

37
PoolingEquilibrium

• Market cannot tell the firms apart uses p
• Good Firms old equity
• Why am I not bothering to calculate bad firms
  equity value?

38
Signaling

• Good firm decides to issue only equity as a
  signal. Lets check that if it does this, bad firm
  will find it suboptimal
• If this strategy is credible, all firms that
  issue equity only will get full info a
• Value to old equity of good firm
• If this signal is credible, good firm prefers to
  signal

39
SeparatingEquilibrium

• Can the bad firm imitate?
• If it does it will have to use all equity and
  get
• Value to old equity of bad firm
• If the bad firm does nothing and sits on its
  cash, it just has its cash C3/1.052.85
• Thus the signaling strategy is credible!

40
Intuition

• First example higher tax on dividends than cap
  gains, yet firm rationally chooses to use
  dividends anyways
• Possible explanation for why firms bypass tax
  advantages of cap gains
• Second example riskless debt available and
  carries tax advantage, yet firm rationally
  chooses to raise equity
• Possible explanation (in addition to trade-off
  theory) for why firms use equity
• Signaling strategies required firms to have some
  cash
• Would not have worked if C0
• Possible reason for firms to carry around extra
  cash
• If firms can use cash or riskless debt to pay for
  investments, they do and informational problems
  go away (try redoing either example with
  I5/1.05max value of riskless debt). If they
  cannot, they resort to equity
• This is an example of pecking order theory

41
Pecking Order

• Above examples showcase a feature of pecking
  order theory, but are signaling examples
• Canonical example of pecking order theory (Myers
  and Majluf 1984) is coming up
• Pecking Order Theory Firms can avoid asymmetric
  info problem by using information insensitive
  securities Cash ? Riskless Debt ? Risky Debt ?
  Equity
• Intuition similar to tax economics lump sum
  taxes are non distortionary, income taxes distort
  incentives

42
PeckingOrder

• Again assume 5 discount rate and risk neutrality
• Firm has existing projects in place which pay off
  50 or 150 with ½ probability
• Firm has new project which requires I100 and has
  NPV 10 in firms bad state or 20 in good state
• When firms existing projects worth 150, new
  project needs 100, pays 120
• NPVgt0 so project should always be taken

43
Full Info

• Suppose firm commits to invest in this project
  and raises equity to do so
• The value of old equity in the bad state is
• (1-a)(50110)81.86
• The value of old equity in the good state is
• (1-a)(150120)138.14
• The present value of old equity is
• This is just as it should be Old Equity New
  Equity 100110/1.05 204.76 Total Value

44
AsymmetricInfo

• Suppose old equity knows the state of the world
  but investors do not
• Note, if state of the world is good, equity gets
  150/1.05 by doing nothing, but only 138.14/1.05
  by raising funds and taking on project
• If state of the world is bad, equity gets 50/1.05
  by doing nothing, 81.86/1.05 by raising funds and
  taking on project
• Old equity will only raise funds and take on
  project in the bad state

45
RationalOutsiders

• But outsiders are not idiots!!!
• If only bad projects are taken, under current
  a105/21549 they will get
• a(50110)/1.0578.14lt100
• They will ask for another, fair a
• 100a(50110)/1.05 ?
• a105/16066
• The value of old equity if firm turns out to be
  bad is (1-a)(50110)/1.0552.38
• The value of old equity if firm turns out to be
  good is 150/1.05142.86

46
ValueLost

• Think back to time zero, before insiders know the
  state
• We solved earlier that value of old shareholders
  if they can precommit to invest in all NPVgt0
  investments is 104.76
• However we know firm cannot precommit and will
  only invest in bad state, ex ante their equity is
  worth
• .552.38.5142.8697.62
• Investors losing from asymmetric info!

47
Intuition

• Even though project has positive NPV in either
  state, the value of the in place assets is so
  high in the good state, insiders do not want to
  share that with new shareholders
• Equity is like profit sharing. As in signaling
  examples, firm only willing to share profit (sell
  equity) when profit is low
• Sometimes firm overvalued, sometimes undervalued.
  When undervalued firm refuses to issue equity and
  so passes up valuable projects
• What if firm had financial slack (carried cash)
  for such occasions?

48
FinancialSlack

• Same example but firm has C100
• If firm pays this cash out as dividend before
  project available, it places itself in same
  situation as in previous example. Total value of
  shareholders is 10097.62197.62
• If firm uses this cash to invest in project in
  either state of world value of shareholders is
  .5(50110).5(150120)/1.05204.76!
• Since NPVgt0 in both states, firm will not choose
  strategy of only investing in one state
• By having financial slack firm gets around
  asymmetric info problem!

49
PeckingOrder

• If firm had private line to its shareholders
  announcing NPVgt0, it could raise money from them
  only (ie negative dividend to create cash). Then
  they would not have to share profits to outsiders
• If firm could issue enough safe debt to finance
  investment, it could avoid the problem since this
  is just like raising C
• In more general model Myers and Majluf show that
  risky debt alleviates, but does not fully solve
  problem
• If firm cannot do any of above and raises equity,
  its stock price falls
• ex ante price includes both states
  .552.38.5142.8697.62
• ex post price reveals bad state 52.38
• soon will see similar pattern in data

50
Pecking Order

• Issuing equity signals bad news about firms
  prospects
• Firms should use safe securities (which reveal no
  info) such as cash, safe debt first. If they
  still need to raise more money, they should move
  up the pecking order risky debt, equity

51
AmbiguousSignals

• Manager increases (decreases) personal share of
  equity. Is this a good (bad) signal about firm?
  Maybe decreases to diversify?
• Firm cuts dividends to raise investment. Is this
  a good thing or signal about shortage of CF?
• Is firm making decision to maximize long term
  value of firm or short term stock price? Is there
  a difference?

52
MarketTiming

• Firms can shift reported income forward or
  backward (ie use different depreciation
  calculation)
• Firms make earnings seem higher before IPO and
  SEO (Teoh, Welch, Wong 1998)
• Firms make earnings seem lower before union
  negotiations (Liberty and Zimmerman 1986)
• SEOs coincide with high valuations (ie M/B, P/E)
  Taggart (1977), Aquith and Mullins (1986),
  Hovakimian, et al (2001)
• IPOs coincide with high valuations
• Loughran, et al (1994), Pagano, et al (1998)

53
QuickQuestion

• How would you test trade off theory? Market
  timing? Pecking order?

54
QuickQuestion

• Trade-off
• Do firms with higher probability of distress
  (high vol) or higher cost have lower cap
  structure?
• Identify optimal cap str. (ie by industry
  average). Do firms move towards this?
• Market Timing
• Identify over (under) valued firms. Do these
  firms issue equity (debt)?
• High recent stock returns ? overvalued? Maybe
  this just means they have lots of investment
  opportunities
• Use fundamentals to identify value?
• Pecking Order
• Are firms with higher info asymmetry issue more
  likely to carry cash? Issue debt?
• Do firms with more investment opportunities carry
  more cash?
• Price drop after equity issues?

55
The Financing Decision Market Timing?

• 2/3 CFOs "amount by which our stock is
  undervalued or overvalued was an important or
  very important consideration" in issuing equity
• Graham and Harvey (2001)
• Firms raise debt and repurchase equity when they
  have less debt than similar firms (target ratio,
  trade-off) raise equity after stock price
  increases (overvalued, market timing)
• Hovakimian, Opler, Titman (2001)
• Other studies suggesting SEOs coincide with high
  valuations (stock run up, M/B, P/E)
• Taggart (1977), March (1982), Korajczyk, Lucas,
  McDonald (1991), Jung, Kim, Stulz (1994)
• East Asia IPO volume positively correlated w/
  stock market level (mkt val high ? IPO). High IPO
  volume predicts lower long run returns (though
  not significant). This is consistent with market
  timing
• Loughran, Ritter, Rydqvist (1994)
• Likelihood of IPO positively related to M/B in
  industry IPO followed by abnormal reduction in
  profit, new equity cap raised not used to finance
  inv. but to reduce lev.
• Pagano, Panetta, Zingales (1998)
• Value stocks (undervalued) more likely to
  repurchase shares
• Ikenberry, Lakonishok, Vermaelen (1995)
• Issues more likely when stock has done well
  relative to market
• Asquith and Mullins (1986)

56
The Financing Decision Pecking Order?

• Equity issued when market has most info (after
  announcements)
• Korajczyk, Lucas, McDonald (1991)
• Fazzari, Hubbard, Petersen (1988) separate firms
  into three groups based on their dividend payout
  ratio (D/P) claim low D/P ? Firms financially
  constrained. For each group they regress
  I/Kab1Qb2(CF/K) where Q is a measure of
  investment opportunities
• Positive and significant b1 for unconstrained
  firms, insignificant b1 for constrained firms ?
  firms that are financially constrained are less
  likely to invest even when good opportunities
  exist ? Why not? Cant get cheap funds
• They find larger b2 for constrained firms ? firms
  that are constrained but have free cash are more
  likely to undertake good investment since less
  dependent on raising funds

57
The Financing Decision Pecking Order?
58
The Financing Decision Pecking Order?

• Pecking Order would predict instantaneous drop
  after announcement, as in table above
• Reaction to equity offerings -2.7 over 2 days
• Asquith and Mullins (1986)
• Stocks react positively to share repurchases too
• Dann (1981)
• Firms w/ investment opportunities more likely to
  issue equity than w/out. when these firms issue
  equity, the negative reaction of stock is smaller
  than avg
• Jung, Kim, Stulz (1994)

59
The Financing Decision Market Timing?

• But prolonged reaction more consistent with
  timing and misinformed investors than pecking
  order and efficient markets .
• Although there are possible rational risk based
  explanations
• Equity issuers have low subsequent returns after
  SEO
• Sigler (1964), Ritter (1991), Loughran and Ritter
  (1995), Speiss and Affleck-Graves (1995), Brav
  and Gompers (1997)
• Abnormal return for repurchases is 12.1/ann and
  45.3 for value stocks
• Ikenberry, Lakonishok, Vermaelen (1995)

60
The Financing Decision Market Timing?

• Baker and Wurgler (2001) Capital structure is
  the cumulative outcome of attempts to time the
  equity market
• Not pecking order or trade-off theory
• Low leverage firms are those that raised funds
  when market valuations (M/B) were high
• Therefore they raised equity
• High leverage firms are those that raised funds
  when market valuations (M/B) were low
• Therefore they raised debt
• Regress leverage on weighted average of past M/B
  ratios, find strong negative relationship
• This is very persistent, lagged valuations have
  half life of 10 years so basically completely
  explain capital structure

61
Costly Signaling

• Remember dividends are costly way to pay out
  (tax reasons), but may act as a signal
• Stock price increases 2 after div increase
  announcement
• Aharony and Swary (1980)
• The increase is even bigger for firms who paid no
  dividend
• Michaely, Thaler, Womack (1995)
• Many firms issue equity and pay dividends at same
  time. This seems puzzling they could reduce
  dividend if they need cash for investment rather
  than issue equity
• 40 of all issuances made by firms with regular
  payout policies, 20 could have been avoided by
  eliminating dividend. The post-SEO negative
  return after is less severe by these types of
  firms
• Weld (2008)
• Long and Litzengerger (1989) Separated firms into
  4 groups
• div incr div
  decrease
• high M/B .003 -.003
• low M/B .008 -.027
• This suggests market reacts rationally,
  punishes firms with low investment opportunities
  who cut dividends, reward low investment
  opportunity firms who raise dividends

62
Conclusion

• Firms are better off raising equity when they are
  overvalued and debt when undervalued
• Rational market anticipates this ? Good firms
  want to let the market know they are good ?
  Pecking order
• Costly signaling allows Cream to rise to the
  top
• Empirical evidence for pecking order is mixed
• Lots of empirical support for market timing