11

A Dynamic Optimal Capital Structure Model

with Costly Adjustment Mechanisms

Hsien-Hsing, Liao
Yi-Hsuan, Tung
Tsung-kang Chen

Department of Finance
National Taiwan University
2008.12.12
Abstract
• In this study, we extend Goldstein, Ju and Leland (2001) to a
cash flow-based model with costly adjustment mechanisms
suggested by Leary and Roberts (2005) and include a fixed
and a proportional cost elements in the mechanisms.
• Main Results:
▫ Costs of capital structure adjustments, including a fixed cost element
and a proportional one, have significant effect on the frequency and
size of adjustments in capital structure by a firm.
▫ Our model is able to consider multiple types of adjustments, such as
issuing and repurchasing either debt or equity.
▫ The numerical results of simulation analysis of our model are
consistent with those expected in literature and intuition.
2
 3

A Dynamic Optimal Capital Structure Model with Costly

Adjustment Mechanisms

Agenda
• Introduction
• Description of the model
• Dynamic Capital Structure Strategy
• Numerical Results
• Conclusion

3
 4

Introduction
MM
Firm value Tax benefit
Financial
distress cost

Optimal leverage
Leverage

• A survey on chief financial officers by Graham and Harvey

(2001, JFE) reports that over 80% of the respondents employ
varying degrees of target debt ratios in their practice.
4
 5

Introduction
• There are substantial debates on the idea of a target
debt ratio:
 Fama and French (2002, RFS) suggest that firms seem slowly to
rebalance their leverage to their long-run mean or optimal level.
 Baker and Wurgler (2002, JF) document that current capital
structure is strongly related to historical market values and firm
times equity issuances with high market valuations and has
persistent effects on capital structure.
 Welch (2004, JPE) observes that stock price changes have a
strong effect on market leverage ratios.
5
 6

Introduction
• Other literatures have disparate estimated results:
 Flannery and Rangan (2005, JFE) find that firms do target a
long run capital structure and that a typical firm converges
toward its long-run target at a rate of more than 30% per year.
 Kayhan and Titman (2007, JFE) indicate that after controlling
for the changes in stock prices and other timing and pecking
order effects, changes in debt ratios are still partially explained
by movements towards a target debt ratio.
6
 7

Introduction
• There are growing literature that considers adjustment
costs to explain the relatively slow movement towards
optimal leverage.
 Fischer, Heinkel and Zechner (1989, JF) develop a model of
dynamic capital structure choice in the presence of
recapitalization costs. They find that even small recapitalization
costs lead to wide swings in a firm's debt ratio over time.
7
 8

Introduction
 Leary and Roberts (2005, JF) shows that the presence of
adjustment costs has significant implications for corporate
financial policy.
◦ They reexamine the conclusions of Baker and Wurgler (2002, JF)
and find that the persistence revealed by their empirical tests is
more likely due to adjustment costs.
◦ They also find that the effect of Baker and Wurgler’s key market
timing variable on leverage attenuates significantly as adjustment
costs decline, proving that adjustment costs appear to dictate the
speed at which firms respond to leverage shocks.
8
 9

Introduction
• These studies ignore the possibility that economic
shocks may cause their optimal capital structure to
change over time.
• There are several theoretical dynamic capital
structure models developed such as
• Fischer, Heinkel and Zechner (1989, JF);
• Goldstein, Ju, and Leland (2001, JB);
• Titman and Tsyplakov (2007, wp).

9
 10

Introduction
• Based on Goldstein, Ju, and Leland (2001, JB), we develop
a dynamic capital structure model that allows us to
determine the value of firm, value of debt, value of
bankruptcy, optimal coupon and leverage
endogenously, and to quantify the benefits and costs
associated with both movements towards and away
from optimal capital structure.
10
 11

Description of the model

• The model developed in this paper follows the EBIT-based
model of dynamic capital structure (Goldstein, Ju and Leland
(2001)).

• Here we consider a simple model of firm dynamics. The firm

produces a cash flow is specified by the stochastic process
d p
  p dt   dZ
(1) 
p
where the growth rate, , 
and the volatility of cash flow, , are constants. 11
 12

Description of the model

• Both the risk premium, θ, and the risk-free rate, r, supported
by this firm are constants.
• It is well known that any asset of the firm can be priced by
discounting expected cash flows under the risk-neutral
measure. In particular, the value of the claim to the entire
cash flow, is
 t
V (t )  E (  e  s ds ) 
Q  rt

r
t
t (2)

where μ = μP - θσ is the risk-neutral drift of the cash flow rate

under risk neutral measure. 12
 13

Description of the model

• Here, we refer to V as the value of the firm. Since r and μ are
constants, both V and δ share the same dynamics:
dV
  dt   dZ Q (4)
V
• We assume a simple tax structure that includes personal as well
as corporate taxes. Interest payments to the investors are taxed
at a personal rate τi , effective dividends are taxed at τd , and
corporate profits are taxed at τc , with full loss offset provisions.
13
 14

Description of the model

• Considering a debtless firm with current firm value V0 .
Rather, both equity and government have a claim to the firm’s
cash flow.
• Then, the current firm value is divided between equity and
government as E  (1   )V
eff 0
(6)
G   eff V0
(1   eff )  (1   d )(1   c ) (7)
where the effective tax rate is
14
 15

Description of the model

• The formulas are as following expressions:

C1
D1  D  V0 , C1 ,VB (C1 )  (1   i )  1  PB (V0 , C1 )  (1   )(1   eff )VB PB (V0 , C1 )
r

 C 
E1  E  V0 , C1 ,VB (C1 )   (1   eff ) V0  VB (C1 ) PB (V0 , C1 )  1  1  PB (V0 , C1 )  
 r 
C0
D0  D  V0 , C0 ,VB (C0 )   (1   i )  1  PB (V0 , C0 )  (1   )(1   eff )VB PB (V0 , C0 )
r
 C 
E0  E  V0 , C0 ,VB (C0 )   (1   eff ) V0  VB (C0 ) PB (V0 , C0 )  0  1  PB (V0 , C0 )  
 r 
15
 16

Description of the model

C. Adjustment cost function
• According to Leary and Roberts (2005)’s paper

16
 17

Description of the model

• We consider that the adjustment cost of rebalancing the capital
structure includes two parts, the fixed cost and the
proportional cost.
• The fixed cost is relative to
size of current asset.
• The proportional cost is
relative to the amount of
issuance or repurchase.

17
 18

Firm value

Description of the model

• The adjustment cost function (AC) : L L L* L* L* D/V
AC  ( F D0  F E0 )
L0  L1 k
 [ P ( D1  D0 )   ( E1  E0 )]( )
L0  L*

(28)
D0 D1 * D*
where L0  , L1  , L  *
0 1 
 , , , , k
F F P P

and are coefficient of adjustment cost, which

depend on firm specific effect and industrial cyclical.
18
 19

Dynamic Capital Structure Strategy

• Our objective is to find out the coupon level, C1, which
maximizes the firm benefit after recapitalization minus
adjustment cost.
Max    1  0   AC , 0 
C1
(29)
where ν1 is the asset value after adjustment and ν0 is the asset
value before adjustment, as follows:
 1  D1  E1  D  V0 , C1 ,VB (C1 )   E  V0 , C1 ,VB (C1 ) 
 0  D0  E0  D  V0 , C0 ,VB (C0 )   E  V0 , C0 ,VB (C0 ) 
19
 20

Dynamic Capital Structure Strategy

• There are two situations that we choose to recapitalize:
1. When the firm value touch to the default value.
2. When the benefit of recapitalization is larger than cost.

• We calculate the firm value, V (eq.[2]), and determine whether

it is larger than the bankruptcy level, VB, at each period.
◦ If V<VB, adjust to optimal coupon level. (eq. [27])
◦ If V>VB, adjust to coupon level C1. (eq.[29])
20
 21

Numerical Results Table 1: The base case parameter values

Parameters and Variables Values
• We simulate 1000
σ, the volatility of cash flow 0.10
random paths for the μ, the growth rate of cash flow 3.5%
cash flow ten years δ, the initial cash flow at period 0 100
r, risk-free rate 4.5%
quarterly, which
q, the risk premium 3.0%
randomly generates τc, the corporate profits tax rate 0.35
the information τd, the dividends tax rate 0.20
τi, the personal tax rate 0.35
embedded in the α, the bankruptcy cost 0.5
various periods. εF, the fixed cost of debt 0.1%
ωF, the fixed cost of equity 0.2%
εP, the proportional cost of debt 1.0%
ωP, the proportional cost of equity 5.0%
k, the adjustment exponent 1 21
 22

Numerical Results Figure 1: Simulated leverage dynamics.

22
 23

Numerical Results Figure 2: The leverage under

different volatility of cash flow.

23
 24

Numerical Results Figure 3: The dynamics leverage and the

size of adjustment coupon. (σ=0.05)

24
 25

Numerical Results Figure 5: The dynamics leverage and the

size of adjustment coupon. (σ=0.15)

25
 26

Numerical Results Figure 6: The leverage under different

risk-free rate environment.

26
 27

Numerical Results Figure 7: The dynamics leverage and the

size of adjustment coupon. (r=6.5%)

27
 28

Numerical Results Figure 9 The dynamics leverage and the

size of adjustment coupon. (r=4.5%)

28
 29

Numerical Results Figure 10: The leverage under different

bankruptcy cost scenarios.

29
 30

Numerical Results Figure 11: The dynamics leverage and the

size of adjustment coupon. (α=0.15)

30
 31

Numerical ResultsFigure 13: The dynamics leverage and the

size of adjustment coupon. (α=0.50)

31
 32

Numerical Results Figure 14: The leverage under different

structure of fixed cost.

32
 33

Numerical Results Figure 15: The leverage under different

structure of proportional cost.

33
 34

Numerical Results Figure 16: The leverage and size of

adjustment coupon under the fixed
cost of adjustment.

34
 35

Numerical Results Figure 17: The leverage and size of

adjustment coupon under the
proportional cost of adjustment.

35
 36

Numerical Results Figure 18: The leverage and size of

adjustment coupon under the large
fixed cost of adjustment.

36
 37

Numerical Results Figure 19: The leverage and size of

adjustment coupon under the large
proportional cost of adjustment.

37
 38

Conclusion
• We modify a EBIT-based Model proposed by Goldstein, Ju,
and Leland(2001), and develop dynamic capital structure
model with costly adjustment mechanisms.
• We consider costs of capital adjustments, including a fixed
cost element and a proportional one, which have significant
effect on the frequency and size of adjustments in capital
structure by a firm.
• Our model is able to consider multiple types of adjustments,
such as issuing and repurchasing either debt or equity.
38
 39

Conclusion
• Based on an exogenous cash flow process, the model can
endogenously determine the firm value and the claim value of
firm’s derivative security under optimal debt level.
• A firm can alter the variables of the model based on industry and
firm specific properties to determine if doing a recapitalization at
any point in time.
• Finally, the numerical results of simulation analysis of our model
are consistent with those expected in literature and intuition.
39
 40

Reference
[1] Baker, Malcolm, and Jeffrey Wurgler, 2002, “Market timing and capital structure”,
Journal of Finance 57, 1–30.
[2] Fama, E. F. and K. R. French, 2002, “Testing trade-off and pecking order
predictions about dividends and debt”, The Review of Financial Studies 15(1): 1-33.
[3] Fischer, E., Heinkel, R., and J. Zechner, 1989a, “Dynamic Capital Structure Choice:
Theory and Tests”, Journal of Finance 44, 19-40.
[4] Flannery M. and K. Rangan, 2006, “Partial Adjustment toward Target Capital
Structures”, Journal of Financial Economics 79, 469-506.
[5] Goldstein, R., Ju, N., and H. Leland, 2001,“ An EBIT Based Model of Dynamic
Capital Structure”, Journal of Business 74, 483-512.
[6] Graham, J., and Harvey C., 2001, “The Theory and Practice of Corporate Finance:
Evidence from the Field”, Journal of Financial Economics 60, 187-243.
[7] Hovakimian, A., Opler T, and S. Titman, 2001, “The Debt-Equity Choice”, Journal
of Financial And Quantitative Analysis, 36 (1), 1-24.
[8] Kayhan A,. and S. Titman, 2007,“ Firms’ Histories and Their Capital Structure”,
Journal of Financial Economics 83, 1-32.
40
 41

Reference
[9] Leary M., and M. Roberts, 2005, “Do Firms Rebalance Their Capital Structures?”,
Journal of Finance ,2575-2619.
[10] Leland,H.E.,1994, “Corporate Debt Value, Bond Covenants, and Optimal Capital
Structure”, Journal of Finance,1213-1252.
[11] Leland, H. E. and K. B., Toft, 1996, “Optimal Capital Structure, Endogenous
Bankruptcy, and the Term Structure of Credit Spreads”, Journal of Finance 51, 987-1019.
[12] Leland, H.E., 1998, “Agency Costs, Risk Management, and Capital Structure”, Journal
of Finance 53, 1213-1243.
[13] Miller, M.H. 1977 “Debt and Taxes,” Journal of Finance 32, No.2, 261-275.
[14] Modigliani, Franco, and Merton H. Miller, 1963, “Corporate income taxes and the cost
of capital: A correction”, American Economic Review, 53, 433-443.
[15] Sheridan Titman and Sergey Tsyplakov, 2007, “A Dynamic Model of Optimal Capital
Structure”, working paper.
[16] Welch, Ivo, 2004, “Capital structure and stock returns”, Journal of Political Economy
112, 106–131.
41
 42

Description of the model

• Consider an otherwise identical firm whose management
decides to choose a static leverage level that will maximize
the wealth of current equity holders.
• We assume a firm will issue a perpetual bond, promising a
constant coupon payment C to debt holders as long as the firm
remains solvent.
• We define the firm chooses the default threshold as VB. If the
firm value reaches VB , then an amount αVB will be lost due to
bankruptcy costs.
42
 43

Description of the model

• In general, any claim of firm’s derivative security must satisfy
the partial differential equation (PDE) as follows:
2 2
VFV  V FVV  Ft  P  rF (8)
2
where P is the cash flow.
• Because of the issuance of perpetual debt, all claims will be
time-independent. Thus the PDE reduces to an ordinary
differential equation:
2 2
VFV  V FVV  P  rF  0 (9)
2
43
 44

Description of the model

• The solution is
y x
F  A0  AV
1  A2V (10)
where
  2  
x  1       2  2
    2r   0
  2  2   2  
     

  
1  2   2  2
y         2r   0
 2
   2   2  
    
• and A0, A1 and A2 are constants, determined by boundary
conditions.
• Note that x >0, while y <0, so A1 equals zero for all claims of
interest. 44
 45

Description of the model

• Before proceeding, it is convenient to define pB (V) as the present
value of a claim that pays $1 contingent on firm value reaching
VB .
y x
pB (V )  0  AV
1  A2V (11)
• According to the boundary conditions
lim pB (V )  0 lim pB (V )  1
V  V VB

• Giving x
V  (12)
PB (V )   
 VB  45
 46

Description of the model

• We define the value of this claim as Vsolv (V) as long as firm value
remained above VB.
y x
Vsolv  V  AV
1  A2V (13)
• Taking into account the boundary conditions:

lim Vsolv (V )  V lim Vsolv (V )  0

V  V VB

• We find that
(14)
Vsolv  V  VB PB (V )
46
 47

Description of the model

• As long as the firm remains solvent, the coupon payment is C
before the recapitalization. Thus, the value of the claim to interest
payments during continued operations, Vint is
C y x
Vint   AV
1  A2V (15)
r
• The boundary conditions are
C lim Vint  0
lim Vint  V VB
V  r
• We obtained (16)
C
Vint   1  PB (V ) 
r 47
,
48

Description of the model

• The present value of the default claim Vdef (V) can be written as
y x
Vdef  0  AV
1  A2V (17)
with the boundary condition:

lim Vdef  0
V 
lim Vdef  V
V VB
• We get
(18)
Vdef  VB pB (V )
48
 49

Vsolv  V  VB PB (V )
Description of the model Vdef  VB pB (V )
• Note that the sum of the present value of the claim to
funds during solvency (eq. [14]) and the present value of
the claim to funds in bankruptcy (eq. [18]) is equal to the
present value of firm, V.
• Hence, value is neither created nor destroyed by changes in
the capital structure; rather, it is only redistributed among the
claimants. This invariance result is consistent with the “pie”
model of Modigliani and Miller (1958), except that in this
framework the claims of government and bankruptcy costs
are also part of that pie. 49
 50

Description of the model

• Separating the value of continuing operations and default claim between
equity, debt, government, and bankruptcy costs as follows:
(19)

E (V )  (1   eff )(Vsolv  Vint ) (20)

D(V )  (1   i )Vint  (1   )(1   eff )Vdef
(21)
G (V )   eff (Vsolv  Vint )   iVint  (1   ) eff Vdef
(22)

BC (V )  Vdef
50
 51

Description of the model

A. Default Level
• When the firm has the financial distress, which means the cash
is less the promised coupon payment to the debtholders.
• We can obtained the bankruptcy level VB by invoking the
smoothing-pasting condition
E
(23) 0
V V VB
• In poor situation, the firm goes into bankruptcy, comparing
with the debt holders get the residual firm value and the equity
holders receive nothing. 51
 52

Description of the model

• Solving, we find
C
VB (C )  
r
(24)
 x 
where   
 x 1

• Observed that the firm value at which bankruptcy occurs

a)is proportional to the coupon, C ;
b)is independent of the current firm value, V ;
c)decreases as the risk-free interest rate, r, rises. 52
 53

Description of the model

B. Optimal Coupon level
• The objective of management is to maximize shareholder
wealth. That is, current equity holders receive fair value for
the debt claim sold.

0
(26)
C V V0

where  (V )  D (V )  E (V )

53
 54

Description of the model

• We find that the optimal coupon level chosen when current
firm value is V0, is
1
r  1  A   x
C  V0 ( ) 
*
 
  1  x  A  B  
(27)

where
A   eff   i
B   (1   eff )

54