Testing Pecking Order Theory and Trade-off Theory
A System Dynamics Approach
Abstract
The fundamental objective of this paper is to present a dynamic framework to test the two
competing theories; the Pecking Order Theory (POT) and the Trade-off Theory (TOT); that
explain the capital structure behavior of firms. For this purpose we use System Dynamics
(SD) method to develop a generic simulation model of a manufacturing firm based on
generally accepted accounting principles. We model the capital structure decision
conforming to POT and TOT to test the two competing theories, in isolation and in
combination. The firms may pursue POT or TOT for their capital structure decision, but it is
generally agreed that while doing so their prime objective is to maximize the firm value.
Hence we presume that the managers stick to the core objective of firm value maximization.
Literature generally suggests the two competing theories as substitutes. We, however,
demonstrate the firms following synergy of the two theories would outperform the firms
following two theories independently in their pursuit of firm value.
Keywords — Capital structure, pecking order, trade-off theory, system dynamics.
1, Introduction
The complexity and strategic importance of long term capital structure behavior of the firms
has resulted into a voluminous debate in corporate finance literature. Pecking Order Theory
(POT) and Trade-off Theory (TOT) are two such competing and influential explanations.
POT, while explaining corporate leverage behavior of the firms, suggests that there is no
well-defined target capital structure rather asymmetric information between the firm and the
market creates a hierarchy of costs in the use of external financing which is broadly common
to all firms and the choice of debt or external equity is a partial function of management's
view of the firm's future prospects. The firms prefer internal to external financing not only to
avoid cost but also to avoid attention by not going to financial markets in view of asymmetric
information. However, if external financing is a must the firms prefer debt over equity
because of lower information costs associated with debt (Myers 1984).
The observations of Myers (1984) are contrary to the TOT, the other competing theory of
firms’ leverage behavior. The proponents of TOT suggest that the firms pursue an optimal
capital structure by evaluating the costs and benefits of the additional financing. For a
comprehensive literature review of both the theories please see (Harris and Raviv 1991).
Corporate finance literature presents three main methods to test the two theories: empirical
evidence, interview or survey, and model based approach. Empirical evidences from various
contexts are mixed and inconclusive (Graham and Harvey 2001; Prasad, Green and Murinde
2001; Fama and French 2002). Some studies support POT (Baskin 1989; Allen 1993;
Adedeji 1998; Shyam-Sunder and C. Myers 1999; Tong and Green 2005; Qureshi 2009)
while others do not (Brennan and Kraus 1987; Vilasuso and Minkler 2001). This indicates
that the outcomes of empirical study may heavily depend on its setting. Likewise, interview
or survey may also assess the expected leverage behavior of the firms’ policy makers in a
given setting. Instead of putting forward some empirical evidence or carrying out interview
or survey and hence avoiding bias due to setting of the study, we develop a generic
simulation model of a manufacturing firm (hereinafter ‘the firm’) using System Dynamics
method based on generally accepted accounting principles, and model the capital structure
decision conforming to POT and TOT to test the two competing theories in isolation and in
combination. The firms may pursue POT or TOT for their capital structure decision, but it is
generally agreed that while doing so their prime objective is to maximize the firm value.
Hence, we presume that the managers stick to their core objective of firm value
maximization.
Apart from introduction in this section we organize rest of the paper as follows: section 2
discusses the model, section 3 presents the analysis, section 4 presents conclusions and
policy recommendations and bibliography is at the end.
2. The Model
The POT and TOT have different implications for leverage behavior of the firms, but it is
difficult to adequately distinguish between the two due to complex network of feedbacks
among different variables (Fama and French 2002). Therefore we consider it useful to
represent this network by using SD method’. The Figure 1 depicts the structure of financial
system of our virtual ‘the firm’ which also defines the conceptual framework of the model.
Product ji ti tl
Investment
Market Investments [*——— Decisions
Situation , i*—
a Operating 7 3
Productivity ——~ Performance |-—s Firm Available
Value Resources
* i *
Dividend Equity Financing
Decisions Decisions
Debt
Financial
Risk
Figure 1. The C onceptual Framework
Adopted from: Qureshi (2007, p.27)
Available production capacity of ‘the firm’ will be guided by its expected order rate and this
serves as investment decision of ‘the firm’. It is generally observed that equity holders of low
return firms would like to have firm’s earnings paid as dividends so that they can invest in
' Fora detailed explanation please see http://www.ifi.uib.no/sd/sdinfo.html
high return firm. On the contrary, investors of high return firm would like the firm to retain
the firm’s earnings and reinvest (Lyneis 1988), and as such we model the dividend decision
of ‘the firm’ as a nonlinear function of retum on equity ratio (Figure 2).
Ingu_uiput
|
i
Fels
fe [or
joe
g
[ | [FeunenE=uiy
’ Figure 2. Dividend Decision
We consider that the management of ‘the firm’ will stick to the firm value maximization
construct in its decision making process, financing decision being one such decision.
Initially, ‘she firm’ is indifferent to debt and equity and hence initial debt to assets ratio is 0.5.
However, to model the future financing decision we first calculate net cash flow of each
period (t) as follows:
NCF, = CR, + B, + NE,—NI;—DC,—-Ik— Pi -Ti-Dt eee Eq. |
where NCF; = Net cash flow
CR, =Cash receipts; cash sales plus collection of credit sales
B, =Borrowing
NE, =New Equity
NI, = New Investment
DC; = Direct costs
I, | =Interest payment
P, =Principal payment
T, =Tax payment
Dt =Dividend payment
The NCF; each period is added to the beginning cash balance to give ending cash balance
which is compared against minimum cash balance determined by the cash policy of ‘the
firm’. The difference of minimum cash balance and ending cash balance gives the desired
cash financing (C;) at the end of period t. The Eq. 2 depicts modeling of total desired external
financing (TDEF;) of ‘the firm’.
TDEF; =max(0, NI; + Ppt Cy- NICFy) oo... .eeeeeeeeeeeseeeeeeeeeeeneteees Eq. 2
whereas we model net intemal cash flow (NICF;) of each period, Eq. 3, following
general pattern of a cash flow statement.
NICE, SCR DGe- he- Te- Disyeravosreesreceemnnrerewes Eq. 3
Following Eq. 4 shows modeling of net desired extemal financing (NDEF;) of ‘the firm’
giving first priority to retained earnings (RE;) (Myers 1984).
NDEF; = TDEF; +max(0, RE}).............ceeeeesssseeeeeeneeeeeeeeeeenenes Eq. 4
POT gives second preference to debt and external equity is used as only the last resort. For
this purpose we use the optimization feature of Vensim® grounded in firm value maximization
construct to find out the optimal composition of debt and equity by specifying a reality check
for each stock.
On the other hand, financing decision of ‘the firm’ under TOT, takes into consideration trade-
off between the debt tax shield which accrues due to presence of debt in capital structure and
the bankruptcy costs. Figure 3 depicts non-linear function we assume to depict the effect of
this trade-off on capital structure decision; between normalized interest tax shield and the
debt assets ratio which we take as proxy of bankruptcy costs. The combined effect of these
two will determine debt financing fraction for each period.
rE am
t [tal ; fs al
eons fe Ke pa [ES
0507 0.2807 £ x josié2 [09005
Tw Fee pe
Th rar fas
TE pa pes
Em pee par
Fae pa fs
aE pa
Ta
New a i New 4 i
i jae Pe
Figure 3. Effect of Interest Tax Shield and Bankruptcy C osts on Debt Financing
Fraction
We multiply this fraction with TDEF; (Eq. 2) to determine new borrowing while testing TOT
in isolation. However, when testing TOT in combination of POT we multiply this fraction
with NDEF; (Eq. 4) to determine new borrowing.
As there is no widespread agreement on whether book or market values are more appropriate
for tests of capital structure theory (Baskin 1989; Prasad, Green and Murinde 2001; Tong and
Green 2005), we use net worth per share as proxy for firm value. This keeps the focus of this
study endogenous.
4. Analysis
For the purpose of this study we first simulate this model to see if POT or TOT is more
effective to maximize the value, and second we also depict combined impact of POT and
TOT on value maximization. Generally any effort to model corporate behavior considers
sales as exogenous and takes certain assumptions about it. For this purpose we consider three
scenarios regarding sales; scenariol assumes no growth, scenario 2 assumes 1% growth,
scenario 3 assumes decline with -1% growth. Moreover, we assume the objective function of
‘the firm’ is to bring increasing trend in book value per common stock. Furthermore, we
assume that all other stocks except debt, equity and cash will at least maintain their initial
level. This assumption not only puts a reality check in place but also helps to isolate capital
structure decision and its impact. Under the three scenarios we simulate the model, for POT
and TOT in isolation as well as in combination, by taking capital structure decision of 0%,
20%, 40%, 60% and 80% debt respectively and remaining to be financed through external
equity. This enables us to demonstrate the impact of increasing leverage on corporate
objective of firm value maximization under different theoretical frameworks.
4,1, Scenario 1 (No Growth)
Figure 1 depicts the simulation results under scenario 1 and different assumptions about debt.
The results demonstrate that the two competing theories are at par if ‘the firm’ assumes a low
leverage policy. But TOT proves to be superior to POT if ‘the firm’ gradually increases its
dependence on leverage. However, in all policy options for leverage, except for very high
debt dependence (80% debt) where TOT is a bit better than combination of POT and TOT, a
combination of POT and TOT outperforms the TOT. It is also interesting to note that a low
leverage policy is relatively more useful for firm value maximization objective whatever
capital structure theory, POT or TOT ora combination, the firms may follow.
4,2. Scenario 2 (Growth)
We present the simulation results under scenario 2 in Figure 2 while taking different
assumptions about debt. The results demonstrate that TOT initially outperforms POT with
low debt (20%) but in the long run POT proves to be better. Under the same debt assumption
(20%) TOT initially stands at par with the combination of POT and TOT to achieve firm
value maximization objective but in the long run it loses its strength and gives way to the
combination of POT and TOT. We also observe similar behavior of ‘book value per
common’ under 40% and 60% debt assumption but with 80% debt assumption TOT and the
combination of POT and TOT stand at par to achieve firm value maximization objective.
However, in all the three cases (40%, 60% and 80% debt) POT under performs. Moreover,
Figure 2 depicts that a low leverage policy is relatively more useful for firm value
maximization objective whatever capital structure theory, POT or TOT or a combination, the
firms may follow. A supplementary observation is that higher debt dependence exacerbates
underperformance of POT.
4,3. Scenario 3 (Decline)
Figure 3 presents simulation results under scenario 3 under different assumptions about debt.
The results demonstrate that POT underperforms under all debt levels. But the TOT which is
generally at par with the combination of POT and TOT to achieve firm value maximization
objective remains a bit better than the combination with higher debt (80%). As we observed
in other two scenarios a low leverage policy is relatively more useful for firm value
maximization objective whatever capital structure theory, POT or TOT or a combination, the
firms may follow. But a supplementary observation is that higher debt dependence initially
exacerbates underperformance of TOT and the combination of POT and TOT but the fim
value, ‘book value per common’ being its proxy, bounces back. Such an observed virtual
behavior conforms to the agency theory of debt.
5. Conclusion and policy implications
For the purpose we develop an SD model of ‘the firm’ conforming to the generally accepted
accounting standards and model its financing decision conforming to POT and TOT. Without
considering the firm value, POT explains managerial priority structure of financing sources
to dynamically represent changes in debt levels. And we conform to value maximizing
construct of TOT and instead of a static view we assume a dynamic role of debt level to
define trade-off of interest tax shield and bankruptcy costs to determine future capital
structure. Considering capital structure as a strategic decision in pursuit of value
maximization agenda we observe that generally TOT proves to be better to POT in achieving
the firm value maximization objective. We put forward POT as a complement to TOT rather
than its substitution and demonstrate that a combination of POT and TOT is better to TOT as
well as POT in isolation. The policy implication of this conclusion is that the firms may
determine their financing needs by giving first priority to internal equity, most commonly
observed corporate behavior, and then consider the trade-off of the costs and benefits of debt
to decide the level of debt in their financing choices. Second policy implication that the firms
may generally adhere to low debt policy in pursuit of their firm value maximization agenda is
an outcome of our observation that low leverage policy generally performs better.
References
Adedeji, A., 1998, "Does the Pecking Order Hypothesis Explain the Dividend Payout Ratios
of Firms in the UK?," Journal of Business Finance & Accounting 25, 1127-1155.
Allen, D. E., 1993, "The pecking order hypothesis: Australian evidence," Applied Financial
Economics 3, 101 - 112.
Baskin, J., 1989, "An Empirical- Investigation of the Pecking Order Hypothesis," Financial
Management 18, 26-35.
Brennan, M., and A. Kraus, 1987, "Efficient Financing under Asymmetric Information,"
Journal of Finance 42, 1225-1243.
Fama, E. F., and K. R. French, 2002, "Testing trade-off and pecking order predictions about
dividends and debt," Review of Financial Studies 15, 1-33.
Graham, J. R., and C. R. Harvey, 2001, "The theory and practice of corporate finance:
evidence from the field," Journal of Financial Economics 60, 187-243.
Harris, M., and A. Raviv, 1991, "The Theory of Capital Structure," Journal of Finance 46,
297-355.
Lyneis, J. M., 1988, Corporate Planning and Policy Design - A System Dynamics Approach,
Pugh-Roberts Associates, Inc.
Myers, S. C., 1984, "The Capital Structure Puzzle," Journal of Finance 39, 575-592.
Prasad, S. J., C.J. Green, and V. Murinde, 2001, "Company financing, capital structure, and
ownership: a survey, and implications for developing economies," SUERF Studies 12.
Qureshi, M. A., 2007, "System Dynamics Modelling of Firm Value," J ournal of Modelling in
Management 2, 24-39.
—, 2009, "Does pecking order theory explain leverage behaviour in Pakistan?," Journal of
Financial Economics 19, 1365 — 1370.
Shyam-Sunder, L., and S. C. Myers, 1999, "Testing static tradeoff against pecking order
models of capital structure," Journal of Financial Economics 51, 219-244.
Tong, G. Q., and C. J. Green, 2005, "Pecking order or trade-off hypothesis? Evidence on the
capital structure of Chinese companies," Applied Economics 37, 2179-2189.
Vilasuso, J., and A. Minkler, 2001, "A gency costs, asset specificity, and the capital structure
of the firm," Journal of Economic Behavior & Organization 44, 55-69.
book value per common stock
book value per common stock
0 5 0 6 0 B&B 0 3
‘Time (¥ ear)
book value per common stock: POT&TOT-20%D
book value per common stock TOT-20%D
book value per common stock: POT-206D
book value per common stock Optidata
book value per common stock
50
0 5 1 1 2 2 30 35 40 4550
‘Time (Year)
book vale percommon stack: POTS-TOT-40%D
book vale percommon stack: TOT-40%D
book va percommon stack: POT-40%D
book value percommon stack: Opti data
book value per common stock
0 5 0 6 0 5 0 3 0 6 0 O 5 0 1 20 38 3 % 40 45 50
Time (Yet) Time (Year)
book value percommon stock: POT&TOT-60%D book value per common stock : POT&TOT-B0%D
book value percommon stock: TOT-60%D book value per common stock :TOT-808%D
book value percommon stock: POT-60%D. book value per common stock: POT-BI%D
book value percommon stock: Opti data ‘book value per common stock : Opti data
ook value per common stock book value per common stock
2 2
15 15
1
05
0 0
0 5 0 6 0 B 0 8 0 6 0 O 5 0 5 0 B H 3 4 45 50
Time (Year)
ue percommon stock: POT-208D
ue percommon stock: POT-408D
ue percommon stock: POT-6O8D
book ale percommon stock: POT- AD
book value pr common stock: TOT-20%D
Dook value percommon stock: TOT-40D
book value per common stock: TOT-€0D
book value percommon stock: TOT-€0%D
book value per common stock book value per common stock
2 2
1
0
0 5 b 6
Soave peroomman sock: POT TOT 28D
Soave peroomon sock: POT TOT 08D
Doak ve per oomman sk: POT TOT OAD
Soa te peroomman sock: POTATOT-9RD
Boater oommon sok: TOT-2D
Deore pe oomman sek; TOT 93)
Deo ve per oomman sek: POT.) <<<
Donker oomman sek PO. FD)
Donk per oomman stick pial, $$$
book value per common stock
bork aeons: POTSTOT 9)
oak ear mmon sock: POTTOT 9)
boa ve pe common stock: POTETOT ID
ook vee per conmon sick: POTETOT SA
ook ape common sock: TOT- 9080
bork rear cammon stock POT 6959)
ook ear eammon stk POT 8D)
ok ear cmmon sok: pt ie$ $$$”
Figure 1: Results of Scenario 1 (no growth)
book value per common stock
0 5 0b 6b NM Gb D0 8 0 6 DH 0 5 OW 6 0 6 0 8 0 4 50
Tie (ea Tine (Yea)
‘book value per common stock: POT&TOT-C-20%.D, 4 book value percommon stock: POT&TOT-G-40%D_| ———H——
‘book value per common stock: TOT-G-20%D. 4444444 book value percommon stock : TOT-G-40%D|
‘book value percommon stock: POT-6:20%D), HHH ‘book value per common stock : POT-G-40%D) @@
‘book value percommon stock: Opti data, $$ book vaiue percommon stock: Opti data.
book value per common stock book value per common stock
0 3
Time (eat)
‘book value per common stock: POT&TOT-G-60%D
‘book value per common stock: T0T-G-0%D)
‘book value per common stock: POT-C-60%D
ook percomn stock Opidata,
3350 0 23
Time (Year)
bookvalu percomm stock: POT@TOT:Gé%D.
bookvalue per common stock: TOT-G 84D)
book value per comm stock: POT-G 8),
book valu per comm stock Of dla
3 35
book value per common stock
0 5 0 6 0 GB 3 3% 0 4
Time (Y ear)
book value per common stock :POT-C-20%D
book value per common stock :POT-C-40D
book value per common stock :POT-C-RFD
book value per common stock :POT-C-AsD
book value per common stock
Time (Y eat}
book ale per common stock: POTSTOT-G-208D
bookvalue pr common stock: POT&TOT-G-404D
book vale prcommon stock: POT&TOT-G-604D
book value prcommon stock: POT&TOT-G-0MD
book value per common stock
0 5 0 1 20 2 3 3 40 45 50
Time (Year)
book value percommon stock: TOT-C-20%D
‘book value percommon stock: TOT-C-40%6D
‘book value percommon stock: TOT-G-6%D
‘book value percommon stock: TOT-C-8%D
Figure 2: Results of Scenario 2 (growth)
book value per common stock book value per common stock
2 1
15
1
05 05
0 0
0 5 0 6 20 3 0% %
Tine (Year)
book value percommn stock: POT&TOT-D-2D
book ale percommn stock: TOT-D-20%D
book vahie peroommma stock: POT-D205D
book vai peroommnn stock Optidta
book value per common stock
0 5 0 6 0 8 3 3 4 6
Tine (Year)
book value percommn stock :POTATOT-D-40%D
book value percommn stock: TOT-D-ORD
book value perconmon tock: POT-D-404D
book valu percommn stock: pt data
book value per common stock
02
0 5 0 6 0 3 0 3%
Time (Year)
book vale proonmon stock: POT&TOT DD
book ve proonmon tock: TOT-D-KD
ook vale proonmon stock: POT-D406D
ook vale proommon stock: Opt data
0 5 0 6 0 % 3 3 40 45 50
Tine (Year)
bookvalueperconmon stock: POTSTOT:D-
Iookvaue percomnon sock: TOT-D-€0D
hook value peroomno sock: POT-D-80%D
bookvlue percomnn sock: Opi deta
book value per common stock book value per common stock
1 1
a7 085
a4 oe a7} \
a \
it . 055
02 04
05 0 6 0 Bb DY 8 0 8 HO
Time (Year)
bookvalupercommon tock: POT-D20D).
book value percomnon tock: POT-D40)),
‘book value per common stock: POT-D-6l&4D| HHH
‘book value per common stock: POT-D-8l&D)
book value per common stock
0 5 W 6 0 G 0 %& 0 6 1
Time (Year)
bok value per oommon stack: POT&TOT-D-20%4,
book value per oonmoa stack: POT&TOT-D-40%4)
booktvalue per oonmoa stack: POT&TOT-D-6%4)
book vale peroonmon stack: POT&TOT-D-80%4),
0 5 0 6 0 2% 3 3% 4 4 50
Time (Y ear)
book vaiueperconmon stock: TOT-D-20%),
book vaiue perconmon stock: TOT-D-40%4)
book value perconmon sock: TOT-D-608)D, $$
book vale perconmon sock: TOT-D-80%)
Figure 3: Results of Scenario 3 (decline)
