Show Me the Information:
Board Independence and D&O Insurance
Shih-Chung Chang
Department of Insurance and Financial Management
Takming University of Science and Technology
Jason Yeh
Department of Finance
The Chinese University of Hong Kong
Directors and Insurance as Governance

In the current environment, companies have a strong
incentive to adopt rigorous governance procedures
because those that fail to do so will be unable to
attract top quality directors and will pay a risk premium
in terms of both director compensation and possibly
officer and director liability insurance.
(Cynthia A. Glassman, SEC Commissioner, in a speech delivered to
the National Economists’ Club, April 7, 2003)
10/15/2010
2
Money Matters…

CUHK has a D&O insurance policy:



Coverage limit = HK$20,000,000 (per claim and aggregate)
Deductible = HK$50,000 (per claim)
Though rare, outside directors have paid to settle
shareholders’ claims (Apr. 23, 2007, WSJ):




10/15/2010
Trans union: 10 directors paid $1.35m (1985)
Enron: 10 directors paid $13m (2005, D&O Insurer paid $155m)
WorldCom: 12 directors paid $24.8m (2005, D&O Insurer paid $36m)
Just for Feet: 5 directors paid $41.5m (2007)
3
Literature: Board Effectiveness

Kumar and Sivaramakrishnan (2008, RFS):
More independent board (lower monitoring
efficiency) may actually perform worse.
 Laux (2010, MS): If board oversight is difficult
and costly, higher director liabilities lead to a
lower level of board oversight, and lower
shareholder value.
 Duchin et al. (2010, JFE): the effectiveness of
outside directors depends on the cost of
acquiring information about the firm
10/15/2010
4
Literature: D&O Insurance

Mayers and Smith (1982, JB), Holderness (1990, IRLE): insurers
providing D&O insurance play an important monitoring role, as
they will scrutinize corporate policyholders and thus monitor their
management.

Core (2000, JLE): insurance premiums are higher when boards
are less independent.
Cao and Narayanamoorthy (2005): D&O liability insurance
premiums decline with increased board independence.
Gillan and Panasian (2008) find that propensity of firms to
purchase insurance increases with board independence.




Chalmers, Dann and Harford (2002, JF): Managerial Opportunism
Boyer (2003, 2007): no significant association between D&O
insurance limits or deductibles and board composition.
10/15/2010
5
The Model: Basic Ideas
Contract design can hardly change the
CEO’s behavior
 The board oversees the CEO through
peer monitoring
 D&O insurance improves the information
transparency
 Corporate governance reforms are not
completely effective

10/15/2010
6
Timeline
10/15/2010
7
Actors

Following the modeling strategy of Adams and Ferreira (2007)
and Kumar and Sivaramakrishnan (2008)

The firm value given the CEO’s project
selection:
Vi  (1  pi ) R  pi qD  G,

The CEO:
i  H,L
H
U CEO
 VH  B   [(1  p) R  pqD  G ]  B
L
U CEO
 VL   [ R  G ]

The Board: Weighted function of the firm and
the CEO (I as the weight)
10/15/2010
8
The Monitoring Cost

The board’s monitoring cost is a function of the
monitoring effort, m, and the D&O insurance
coverage, K. Let
1
TC (m, K )   c( K )  m 2
2
c( K )  ce  K

The function is convex in monitoring effort, m.
 The cross-derivative is negative, so the higher
D&O insurance coverage leads to lower
marginal monitoring costs.
10/15/2010
9
Lemma 1
Lemma 1: Given the range of the degree of
information transparency , there exists a1 [0, ]
such that
(a) If 0    1 , then m (K , K )  0 ; Similarly, if 1   ,
then m (K , K )  0.
m*

m
(

,
K
)
(b)
0




Higher K leads
to lower m*
Higher K leads to higher m*
1
10/15/2010

10
Implications


Higher board compensation leads to higher
monitoring efforts.
However, higher D&O insurance coverage can lead
to higher or lower oversight, depending on the
degree of information transparency (λ).


10/15/2010
When λ is low, additional insurance reduces expected
liability payments more than marginal monitoring costs, so
the optimal monitoring effort will be lower.
When λ is high, additional insurance reduces more marginal
monitoring costs than expected liability payments, so the
optimal monitoring effort will increase.
11
Proposition 1
Proposition 1：There exists an unique
Subgame Perfect Equilibrium (SPE),
{(   , K  ), m (   , K  )} ,where (a) an interior
solution  , 0     1 , always exists; (b)
there exist 1 and 2 such that K   0 when
0    1 , and K   0 when 1    2 .
( 2   is the upper limit of the
information transparency.)
10/15/2010
12
Implications

Higher equity ownership by directors leads to
higher monitoring level, thereby increasing the
firm value.
 However, higher D&O insurance coverage can
lead to higher or lower oversight, depending
on the degree of information transparency (λ).


10/15/2010
When λ is low, optimal insurance coverage is zero.
When λ is high, more insurance coverage can lead
to higher degree of optimal monitoring, thereby
improving the firm value.
13
Managerial Opportunism?






Chalmers et al. (2002) use a sample of 72 IPO firms (1992-6) and
find a significant negative relation between the three-year postIPO stock price performance and the insurance coverage
purchase.
According to them, one plausible interpretation is that, like
insider securities transactions, D&O insurance decisions
reveal opportunistic behavior by managers.
However, our model indicates an possible alternative explanation
which is less value judgmental.
Insurance reduces two costs: expected liability payments and
monitoring costs.
The economic trade-off matters!!!
In some cases (low λ firms), insurance does not work, because it
reduces more expected liability payments than monitoring costs,
so the monitoring effort will be lower, so is the firm value.
10/15/2010
14
Proposition 2
Proposition 2：If 1    2 , higher board
independence implies higher optimal
d
 0 , and higher
board’s compensation,
dI
optimal D&O insurance coverage , dK  0 ;
dI



However, If
1 , the optimal D&O

K
insurance coverage is zero,  0 , and
higher board independence only implies
higher optimal board’s compensation, d  0.



dI
10/15/2010
15
Implications

Higher degree of board independence implies
higher optimal board’s compensation
 If the interior solution exists, higher degree of
board independence implies higher optimal
D&O insurance coverage.
 If the optimal insurance coverage is zero, the
degree of board independence has no impact
on the optimal D&O insurance coverage .
10/15/2010
16
Proposition 3
Proposition 3: When the information
transparency parameter is within the range
of 1    2 , the D&O insurance purchase will
increase the board’s monitoring level.
Subsequently, higher board independence will
lead to higher monitoring efforts. On the other
hand, when   1 , a more independent board
will request more insurance, which points to a
lower optimal monitoring effort AND lower firm
value. In such situation, the best strategy is to
purchase zero D&O insurance.
10/15/2010
17
Implications

Board independence affects oversight through three
ways:







dm (  ( I ), K ( I ), I ) m
m d
m dK 




dI
I
 dI
K  dI
 
 
 ?



The direct effect: positive.
The indirect effect via board compensation: positive.
The indirect effect via D&O insurance: varies with the
information transparency.
10/15/2010
18
Proposition 4
Proposition 4: (Firm Value Maximization)
From the perspective of shareholders,
there exists an interval of information
transparency 1    2 , shareholders will be
willing to purchase D&O insurance. In
such situation, the optimal insurance
coverage shows a substitute relationship
with optimal board compensation.
10/15/2010
19
Model Extensions
We treat the D&O insurance premium
rate, litigation risk, and information cost
exogenous variables.
 But how do they affect board
compensation, D&O insurance purchase,
and monitoring level at equilibrium?

10/15/2010
20
Proposition 5
Proposition 5: Higher D&O insurance
premium rate (r) implies higher optimal
d
board’s compensation, dr  0 ; However,
higher D&O insurance premium rate
leads to lower optimal D&O insurance
dK
coverage dr  0 .


10/15/2010
21
Board Independence and D&O Insurance
Premium Rate

If the insurer offers a rate reduction for a more
independent board, would that lead to higher
oversight?
m d
m dK dr
dm[   (r ( I ), I ), K  (r ( I ), I ), I )] m m d  m dK 

(

)



dI
I
  dI
K  dI
  dr
K  dr dI


 



 
Again, it depends!

When λ is low, higher board independence leads to

lower premium rate and perhaps LOWER
monitoring. [The first (+) dominates the second (-).]
When λ is high, higher board independence leads to
lower premium rate and subsequently HIGHER
monitoring. [The second (-) dominates the first (+).]
10/15/2010
22
Proposition 6
Proposition 6: There exists a cut-off value
D̂ , where D̂  K . When D  Dˆ, higher litigation
risk (q) will lead to lower optimal
monitoring level; only when D is large
enough, the higher litigation risk (q) has
the potential to increase the monitoring
level.

10/15/2010
23
Litigation Risk and Board Oversight

Will higher litigation risk lead to higher oversight?
dm (   (q), K  (q), q) m m d  m dK 



dq
q
  dq K  dq
 
 

Only so if it HURTS!

The direct effect is positive only if
ˆ  [ I (1    )  (1  I )(1   )]1 K   K 
D


Only when the out-of-pocket payment is significant
enough will the direct effect surpass the
combination of the other two.
Consist with Laux (2010).
10/15/2010
24
Information Costs and Board
Oversight

Duchin et al. (2010): The effectiveness of outside directors
depends on the cost of acquiring information about the
firm.
 When the cost of acquiring information is low,
performance increases when outsiders are added to the
board, and when the cost of information is high,
performance worsens when outsiders are added to the
board.
 If there exists two different cost structures, given the
optimal D&O insurance coverage K*, due to the
differences in the level of acquiring information: high cost
c1 and low cost c2 , so
1 K   2
TC1 ( K , m )  c1e m
2



We can obtain that m1* < m2*, and
10/15/2010
1 K   2
TC1 ( K , m )  c1e m
2


E[V1 (  , K  )]  E[V2 (   , K  )]
25
Concluding Remarks

We contribute to corporate governance
literature:



10/15/2010
Address the interactions of board independence
and D&O insurance.
Characterize the importance of information
transparency in determining the optimal D&O
insurance coverage
Enrich discussions on board independence and
show that high level of information transparency
helps an independent board achieve higher level of
monitoring efforts.
26