A Growth Type Explanationfor Capital Structure Persistence

Xueping Wu and Chau Kin Au YeungCity University of Hong Kong

Presented at NTU on December 11, 2008, Taiwan

Motivation

• A: Unobserved firm heterogeneities

Road Map1. Existing theories

2. Our main idea A Growth Type Story

3. Empirical: Persistence across growth types leverage ratios important firm fundamentals financing activities

1. Our view about market timing

2. Horse race: growth type vs. market timing

4. Conclusion

Literature• Competing Capital Structure Theories

1. Tradeoff theory

2. Pecking order

3. The market timing argument of Baker and Wurgler (2002)

Can they explain leverage persistence?

1. Tradeoff Theory• Focus on costs & benefits of debt finance Target leverage

Graham and Harvey (2001): Yes in survey

Tradeoff force = adjustment toward target (optimum)

An increase in debt has

Benefits: Tax shield, Disciplining role (for FCF)

Costs:Assets substitution (opportunism), Financial distress, Debt overhang

Tradeoff Theory (Problems)• Persistence = Target leverage? Conceptually, stay where you are ≠ at the target

• Tradeoff force? Titman & Wessels(1988); Rajan & Zingales(1995):

Profitability↑ leverage↓ (the big scar!) Graham (2000): Debt conservatism Fama and French (2002): “Snail” speed of adj. Chang & Dasgupta (2007): Mechanical Mean

Reversion

2. The Myers (1984) Pecking Order• Focus on costs of equity finance Myers and Majluf (1984): adverse selection effect Financial slack is valuable, weakening tradeoff force

(explaining the big scar in tradeoff theory.)

Myers (1984): Pecking order in financing Retained earnings < Debt < Equity (as the last

resort) Asymmetric Information costs

Slow adj., no target (persistently kept away)

The Pecking Order (A Problem)• Equity as last resort? Small growth firms issue a lot of new equity

Rajan & Zingales(1995), Fama and French (2002) and Frank and Goyal (2003)

Why no fear for the AI/adverse selection effect?

The last resort puzzle (the deep wound!)

3. Market Timing • Baker and Wurgler (2002) Capital structure is a result of past attempts to time

the market (especially via equity issues) Market timing is driven by overvaluations (cheap

equity, Stein, 1996).

So equity is not necessarily the last resort Like in Myers (1984): slow adj., no target

3. Market Timing (Debate) • Leary and Roberts (2005): Clustered quick rebalance

• Hovakimian (2005); Kayhan and Titman (2006): Wrong interpretation by Baker and Wurgler

• Lemmon, Roberts & Zender (2008): Leverage persistence: initially determined

• But, Huang & Ritter (2007): Address cheap equity!

Behavior finance: Overvaluation = Cheap new equity

This Paper’s View• MM Theorem:

If capital markets are imperfect, capital structure matters.

• Persistent agency conflicts and AI types across firms (persistent relative agency/AI costs)

Persistently distinct market imperfections across firms

Persistently distinct financing behavior (and capital structure) in response

• A type-determination story for leverage persistence

An Analogy of Type Determination• Baby » Adult:

We observe persistent characteristics, e.g., heavy vs. thin

• Gene determination: diet behavior is gene-driven A heavy kid/adult always eats a lot.

A thin kid/adult cannot eat much.

• The analogy: Gene = Persistent type behind leverage Diet behavior = Financing (incl. market timing)

Asset Type and Asymmetric Info Type(1) Asset Type (Myers, 1977): Tangibles vs. Intangibles(2) AI Type (correlated with Asset Type in reality):

Myers and Majluf (1984)

Type A: more AI about AIP than about growth Type B: more AI about growth than about AIP

Asymmetric Information (AI) arises from

Assets-in-place (AIP) Growth (NPVGO)

AI Types and Financing Patterns

The generalized Myers-Majluf ModelCooney & Kalay (1993) and Wu & Wang (2005)

Dominant Asymmetric Information about AIP

Dominant Asymmetric Information about Growth

The Classic MyMj Framework:

1. Separation of Overvalued Firms from Undervalued Firms at New Equity Issues

2. AI Can Inhibit New Equity Issues Due to Adverse Selection

3. Myers’(1984) Pecking Order in Financing

Generalized MyMj Framework:

1. Not All Growth-oriented Issuers Are Lemons Because Undervalued Firms May Issue. (Pooling)2. An Increase in Asymmetric Information Can Facilitate New Equity Issues.3. Equity Issuers Not Necessarily under Duress

Growth Type and Cost of Capital

A description of agency and asymmetric information costs based on Myers (1977) and the generalized MyMj model

Firm Characteristic Dominant Type of

Asymmetric Information

(AI)

Growth Type

Cost Structure of External

Finance

Market-to-book

Asset Tangibility Debt Equity

Low HighAI about Assets-in-

place Low Growth

(G1) Low High

High High Mixed Mixed Growth (G2)

↓ ↑Low Low Mixed Mixed Growth

(G2)

High Low AI about Growth

High Growth (G3) High Low

A Growth Type Story for Leverage Persistence 1. Lemmon, Roberts & Zender (2008) Initially determined (implying a type story)

2. This paper suggests Growth type gives rise to persistently distinct market

imperfections across firms Growth type preset persistently distinct financing

behavior in responseThus: Growth type explains leverage persistence

parsimoniously. Market Timing is a growth-type driven phenomenon

Test Design: Initial Growth Type

• CompuStat US non-utility & non-financials

• 1971-2005

• Initial value: average of event year 0, 1, 2.

• With breakpoints at medians, two way independent sort on initial market-to-book and initial tangibility

Low High

Low G2 G1(Low Growth)

High G3(High Growth)

G2

Initial Tangibility

Initial

Mar

ket-

to-b

ook

Fig. 1A: Group Mean Book Leverage with the IPO Sample (1971 – 2005)

0.00

0.10

0.20

0.30

0.40

0.50

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

s

G1 G2 G3

Fig. 1B: Group Mean Market Leverage with the IPO Sample (1971 – 2005)

0.00

0.10

0.20

0.30

0.40

0.50

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

s

G1 G2 G3

Fig. 1C: Group Mean Book Leverage with the Full Sample (1971 – 2005)

0.00

0.10

0.20

0.30

0.40

0.50

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

s

G1 G2 G3

Fig. 1D: Group Mean Market Leverage with the Full Sample (1971 – 2005)

0.00

0.10

0.20

0.30

0.40

0.50

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

s

G1 G2 G3

Table 3: Long Lived Effects of Initial Growth Type -- Event Time Cross-sectional OLS

Initial MtB overwhelms updated MtB in the long run. Updated Tang overwhelms initial Tang eventually.

Event

Year

t

Firm

Obs.

Initial

MtB

Initial

Tang MtB t-1 Tang t-1 Profit t-1 LnSize t-1

Ind_Median t-1 DivPayer t-1 Adj. R2

Panel A: Book Leverage

5 7,360 -0.0117 0.2200 -0.1195 0.0203 0.5766 -0.1030 0.191

(-8.49) (18.55) (-8.77) (14.65) (16.21) (-21.07)

-0.0084 0.0505 -0.0080 0.1854 -0.1179 0.0204 0.5214 -0.1058 0.207

T-value (-5.63) (2.31) (-5.49) (8.98) (-8.89) (14.78) (14.55) (-21.82)

10 4,170 -0.0116 0.1570 -0.1510 0.0186 0.4660 -0.0761 0.138

(-6.09) (9.46) (-6.00) (11.03) (9.11) (-11.76)

-0.0108 -0.0188 -0.0052 0.2306 -0.1577 0.0193 0.3880 -0.0821 0.171

(-5.63) (-0.82) (-3.25) (10.84) (-6.09) (11.61) (7.46) (-12.84)

15 2,354 -0.0136 0.0943 -0.1509 0.0191 0.4974 -0.0739 0.134

(-6.07) (4.16) (-4.23) (8.90) (8.37) (-8.57)

-0.0122 -0.0623 -0.0064 0.2438 -0.1672 0.0194 0.3991 -0.0780 0.181

(-5.58) (-2.24) (-2.57) (9.72) (-4.74) (9.38) (6.80) (-9.37)

20 1,448 -0.0117 0.1197 -0.1832 0.0223 0.3368 -0.0929 0.135

(-2.50) (4.05) (-2.65) (8.18) (4.91) (-7.71)

-0.0090 0.0166 -0.0170 0.1467 -0.1614 0.0233 0.2469 -0.0964 0.161

(-2.00) (0.46) (-3.71) (4.60) (-2.14) (8.76) (3.46) (-8.09)

Table 3: Long Lived Effects of Initial Growth Type (cont’d)

Initial growth type (especially via initial MtB) has a strong long lasting effect!

Event

Year

t

Firm

Obs.

Initial

MtB

Initial

Tang MtB t-1 Tang t-1 Profit t-1 LnSize t-1

Ind_Median t-1 DivPayer t-1 Adj. R2

Panel B: Market Leverage

5 7,360 -0.0344 0.2584 -0.1790 0.0253 0.5016 -0.1279 0.287

(-19.95) (18.57) (-12.03) (15.64) (22.03) (-19.82)

-0.0241 0.0728 -0.0262 0.1955 -0.1859 0.0248 0.4243 -0.1271 0.316

(-13.21) (2.98) (-8.14) (8.38) (-13.02) (15.43) (17.96) (-20.04)

10 4,170 -0.0307 0.1994 -0.2545 0.0231 0.4939 -0.0792 0.247

(-13.21) (10.38) (-7.52) (11.49) (16.96) (-9.73)

-0.0258 -0.0191 -0.0257 0.2706 -0.2739 0.0233 0.3962 -0.0837 0.298

(-6.52) (-0.73) (-1.82) (10.94) (-7.67) (11.56) (10.15) (-10.64)

15 2,354 -0.0287 0.1344 -0.2718 0.0187 0.4460 -0.0927 0.201

(-11.39) (4.99) (-6.98) (7.51) (9.66) (-9.38)

-0.0236 -0.0397 -0.0230 0.2545 -0.2970 0.0190 0.3337 -0.0961 0.259

(-5.97) (-1.22) (-1.75) (8.81) (-7.86) (7.92) (6.86) (-10.22)

20 1,448 -0.0256 0.1367 -0.4572 0.0195 0.4359 -0.1133 0.219

(-7.84) (4.04) (-5.78) (6.05) (8.86) (-8.10)

-0.0179 0.0143 -0.0474 0.1514 -0.3787 0.0207 0.3140 -0.1171 0.271

(-4.77) (0.35) (-5.31) (4.18) (-5.27) (6.67) (6.19) (-8.70)

Table 4: How Stable is GT with Updated Tangibility?

The dominant diagonal effect indicates stable growth type.

Transition Matrix from Initial Growth Type, IGT, to GTt

Growth Type with Initial Tangibility replaced by Updated Tangibility

in Event Year t (GTt)

G1 G2 G3

Panel A: Aggregate Average

IGTG1 97% 3% 0%G2 5% 94% 1%G3 0% 5% 95%

Panel B: Simple average

IGTG1 94% 6% 0%G2 3% 94% 3%G3 0% 2% 98%

Fig. 2A: Group Mean Book Leverage with Full Sample in Calendar Time

0.00

0.10

0.20

0.30

0.40

0.50

0.60

1971 1975 1979 1983 1987 1991 1995 1999 2003

Lev

erag

e R

atio

s

G1 G2 G3

Fig. 2B: Group Mean Market Leverage with Full Sample in Calendar Time

0.00

0.10

0.20

0.30

0.40

0.50

0.60

1971 1975 1979 1983 1987 1991 1995 1999 2003

Lev

erag

e R

atio

s

G1 G2 G3

Fig. 3A: Industry-adjusted Book Leverage with Full Sample in Event Time

-0.12

-0.08

-0.04

0.00

0.04

0.08

0.12

0.16

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

min

us I

ndus

try

Med

ian

n

G1 G2 G3

Fig. 3B: Industry-adjusted Market Leverage with Full Sample in Event Time

-0.12

-0.08

-0.04

0.00

0.04

0.08

0.12

0.16

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Lev

erag

e R

atio

min

us I

ndus

try

Med

ian

n

G1 G2 G3

Fig. 3C: Industry-adjusted Book Leverage with Full Sample in Calendar Time

-0.21

-0.18

-0.15

-0.12

-0.09

-0.06

-0.03

0.00

0.03

0.06

0.09

0.12

0.15

0.18

0.21

1971 1975 1979 1983 1987 1991 1995 1999 2003

Lev

erag

e R

atio

min

us I

ndus

try

Med

ian

n

G1 G2 G3

Fig. 3D: Industry-adjusted Market Leverage with Full Sample in Calendar Time

-0.21

-0.18

-0.15

-0.12

-0.09

-0.06

-0.03

0.00

0.03

0.06

0.09

0.12

0.15

0.18

0.21

1971 1975 1979 1983 1987 1991 1995 1999 2003

Lev

erag

e R

atio

min

us I

ndus

try

Med

ian

n

G1 G2 G3

Table 5: Significance of Differences in Leverage Between the Growth Types

Annual Average Number of Firms

Book Leverage Market Leverage

G1 – G2 G2 – G3 G1 – G2 G2 – G3

EventYear G1 G2 G3

t-valuefor

MeanDifference

p-valuefor

MedianDifference

t-valuefor

MeanDifference

p-valuefor

MedianDifference

t-valuefor

MeanDifference

p-valuefor

MedianDifference

t-valuefor

MeanDifference

p-valuefor

MedianDifference

3 to 5 2,373 2,863 2,983 12.74 0.00 24.77 0.00 30.39 0.00 37.08 0.00

6 to 10 1,716 1,884 1,744 9.27 0.00 22.48 0.00 23.08 0.00 33.20 0.00

11 to 15 1,067 1,111 875 11.18 0.00 8.75 0.00 16.71 0.00 16.69 0.00

16 to 20 658 670 441 7.43 0.00 9.51 0.00 10.08 0.00 14.78 0.00

Table 6: Persistent Firm Fundamentalsacross the Growth Types

Growth types are persistently distinct.

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year Market-to-Book Tangibility

3 to 5 0.71 1.25 2.45 -31.14 -36.62 0.67 0.52 0.31 40.07 55.33

6 to 10 0.83 1.22 2.19 -24.17 -32.66 0.64 0.54 0.33 30.60 56.61

11 to 15 0.97 1.29 2.15 -15.15 -21.13 0.60 0.52 0.35 18.16 35.58

16 to 20 1.01 1.32 2.04 -10.06 -12.37 0.58 0.50 0.35 13.42 24.71

Table 6: Persistent Firm Fundamentals (cont’d 1)

• Firm size: G1 > G2 > G3 (Gaps persist.)

• Profitability:G1 > G2 > G3 (stable bumpy)

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year LnSize Profitability

3 to 5 18.86 18.27 17.84 15.02 13.29 13.68 10.13 -1.82 12.05 27.64

6 to 10 19.15 18.59 18.16 14.82 12.61 14.07 12.21 2.19 7.51 26.71

11 to 15 19.44 19.01 18.51 8.62 10.54 12.21 11.76 5.30 1.55 14.48

16 to 20 19.70 19.42 18.83 4.29 8.44 12.37 12.09 7.16 0.82 8.59

Table 6: Persistent Firm Fundamentals (cont’d 2)

• Among profitable firms: G1<G2> G3• Among loss-making firms: G1 > G3 G3: smaller-sized with the largest dispersion in

profits skewed towards losses

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year Profitability > 0 Profitability < 0

3 to 5 14.98 15.72 15.04 -4.09 3.38 -10.49 -17.54 -27.82 5.22 11.22

6 to 10 15.37 15.94 14.98 -3.62 5.34 -10.04 -16.66 -24.74 5.50 8.00

11 to 15 13.88 14.86 14.67 -5.05 0.86 -8.09 -14.05 -23.07 4.67 6.83

16 to 20 13.72 14.67 14.77 -4.07 -0.33 -9.12 -13.23 -21.50 2.55 4.52

Table 6: Persistent Firm Fundamentals (cont’d 3)

• Annual sales growth (SGR): G1 < G2 < G3

G3 firms do deliver the highest sales growth.

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year Asset Growth Rate (%) Sales Growth Rate (%)

3 to 5 9.65 13.46 22.75 -6.07 -9.35 12.36 17.22 30.93 -5.96 -10.77

6 to 10 12.48 13.91 18.41 -2.38 -4.90 12.20 13.91 22.19 -2.93 -7.94

11 to 15 8.78 10.89 16.62 -2.80 -4.97 6.11 9.53 17.55 -4.40 -6.26

16 to 20 7.54 7.43 12.48 0.14 -4.36 7.89 8.17 13.20 -0.31 -3.90

Table 6: Persistent Firm Fundamentals (cont’d 4)

Distinct investment styles:

• Tangible Investment (Capex): G1 > G2 > G3

• Intangible investment (R&D): G1 < G2 < G3 G1 = tangible-growth type; G3 = intangible-growth type

Persistently large R&D for G3 supports its high market to book and high sales growth rate.

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year Tangible Investment or Capext/Assett-1 (%) Intangible Investment or R&Dt/Assett-1 (%)

3 to 5 7.11 7.32 5.54 -1.47 13.99 2.07 4.65 14.48 -21.16 -42.01

6 to 10 7.60 7.00 5.35 4.97 14.82 2.32 4.39 12.58 -19.27 -39.53

11 to 15 7.02 6.41 5.13 4.56 10.11 2.81 4.67 11.32 -13.20 -24.57

16 to 20 6.31 5.87 4.63 3.06 8.63 2.91 4.34 10.17 -9.52 -18.64

Table 6: Persistent Firm Fundamentals (cont’d 5)

• Cash Holdings: G1 < G2 < G3

• %Payers: G1>G2>G3

G1: low cash holdings and typical div payers

G3: high cash holdings and typical div non-payers

EventYear Pooled Mean

t-value forMean Difference Pooled Mean

t-value forMean Difference

G1 G2 G3 G1 - G2 G2 - G3 G1 G2 G3 G1 - G2 G2 - G3

Year Casht/Assett-1 (%) Percentage of Dividend Payers (%)

3 to 5 7.37 14.54 38.38 -20.68 -35.81 56.20 32.60 10.25 21.87 28.17

6 to 10 8.27 13.17 32.35 -14.52 -32.69 64.21 42.12 15.63 21.85 31.69

11 to 15 9.99 13.48 29.16 -7.85 -20.93 65.93 48.06 23.28 13.88 20.92

16 to 20 8.49 11.54 26.13 -7.29 -18.19 66.38 50.97 29.79 9.50 12.85

How about Funding Activities? Fig. 4A: Net Debt Issue by Initial Growth Type

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Net

Deb

t Iss

ue

s

G1

G2

G3

Little differences here

Fig. 4B: Net Equity Issue by Initial Growth Type

• Difference in Net Equity Issues: G1 < G2 << G3 Heavy equity finance is a growth type phenomenon.

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Net

Equ

ity I

ssue

s

G1

G2

G3

Fig. 4C: Change in Retained Earningsby Initial Growth Type

• Difference in RE Changes: G1 > G2 >> G3 (negative!) G3 firms grow largely through equity financing.

-0.12

-0.10

-0.08

-0.06

-0.04

-0.02

0.00

0.02

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Cha

nge

in R

etai

ned

Ear

ning

s

s

G1

G2

G3

For G3, heavy losses, due to the expensing of large R&D investments that pay off slowly, offset new equity issues.

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Net

Equ

ity I

ssue

s

G1

G2

G3

-0.12

-0.10

-0.08

-0.06

-0.04

-0.02

0.00

0.02

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Event Year

Chan

ge

in R

etai

ned

Ear

nin

gs

s

G1

G2

G3

Market Timing (MT)1. Traditional, rational Views (No ex ante under-/over-valuation)

2. Irrational or behavioral view MT is driven purely by market overvaluation

3. Generalized View(Rational, Type-based View)

(1) Traditional Views• Market conditions drive equity issue spikes (MT)

(e.g., Market-to-book, stock price run-ups)

(1a) A decrease in AI of AIP

(Korajczyk, Lucas, McDonald, 1990, 1993; Choe, Masulis, and Nanda, 1993)

(1b) A decrease in the adverse selection effect due to project delay ability (Lucas & McDonald, 1992)

Both are rational views but are silent about σ(growth opportunities)↑.

How about Growth Uncertainty?• Two different views on this observation:

σ(growth)↑ equity issuing costs ↓

(2) Behavioral view Stein (1996), Backer and Wurgler (2002) Exploitable market timing based on Overpricing

& dumb equity investors (Stein, 1996)

(3) The generalized MyMj view Cooney and Kalay (1993); Wu and Wang (2005) Type-based fair market timing

Growth-type-based Fair Market Timing

• Implications of market conditions (MtBt)

Type A: AI mainly about assets-in-place

MtBt ↑ AIt ↓ Equity issuet ↑

Type B: AI mainly about growth opportunities

MtBt ↑ AIt ↑ Equity issuet ↑

“Fair”=Rational expectation is imposed.

Table 7: D and E Issues by IGTFull sample 1971-2005, Pooled OLS with Firm FE

Fair market timing: during market upturn, MtBt ↑

AI about AIP ↓ (good for G1), and AI about growth ↑ (good for G3)

A growth-type-based pecking order in financing

Baker and Wurglar (2002): Low (high) leverage firms tend to raise fund when their valuations are high (low). (Untrue!)

    ΔDebt/Assett   ΔNet Equity/Assett

        Book Value   Market ValueMtBt-1 G1 0.0110 (8.72) 0.0118 (10.00) 0.0287 (13.68)

  G2 0.0068 (9.75) 0.0152 (23.09) 0.0311 (26.68)

  G3 0.0023 (6.83) 0.0161 (50.96) 0.0349 (62.36)Tangt-1 G1 0.0165 (2.90) 0.0158 (2.97) -0.0001 (-0.01)

  G2 0.0062 (1.18) 0.0490 (9.93) 0.0312 (3.57)

  G3 0.0106 (1.77) 0.0953 (17.05) 0.0714 (7.20)

Table 7 (cont’d): D and E Issues by Growth Type

• Profits: A tradeoff force for debt issue at G1 but weak at G3.

• Size: A maturity effect (external finance↓, but on diff. paths)

Taken together: Growth-type determined dynamic external finance

Profitt-1 G1 0.0745 (10.21) -0.0455 (-6.65) -0.0596 (-4.91)

  G2 0.0554 (9.76) -0.0975 (-18.31) -0.1430 (-15.15)

  G3 0.0167 (4.19) -0.1857 (-49.56) -0.2292 (-34.49)LnSizet-1 G1 -0.0112 (-11.12) -0.0108 (-11.39) -0.0158 (-9.43)

  G2 -0.0093 (-10.19) -0.0144 (-16.74) -0.0193 (-12.66)

  G3 -0.0080 (-8.49) -0.0308 (-34.67) -0.0524 (-33.31)Ind_mediant-1 G1 -0.0394 (-2.72) 0.0015 (0.11) -0.0187 (-0.78)

  G2 -0.0876 (-6.29) -0.0089 (-0.68) -0.0858 (-3.71)

  G3 -0.0865 (-5.24) -0.0055 (-0.36) -0.0878 (-3.20)DivPayert-1 G1 0.0253 (13.81) 0.0019 (1.09) -0.0009 (-0.30)

  G2 0.0234 (11.91) 0.0014 (0.74) 0.0023 (0.70)

  G3 0.0154 (5.45) 0.0041 (1.54) 0.0042 (0.88)

Firm FE   Yes Yes Yes

Obs.   76,454 76,454 76,454

R2   0.180 0.512 0.533

Horse Race: Growth Type vs. MT• Baker and Wurgler’s (2002)

For each firm at t:

• es=net equity issue in event year s

• ds=net debt issue in event year s

• Two-Horse Race: (1) B&W MtBefwa, t-1 (Baker-Wurgler MT factor)

versus (2) Initial MtB efwa, t-1: (our growth-type-based variant)

The same weighting scheme but with time-varying annual M/Bs being replaced by initial MtB

It is better than the trailing average MtB, as suggested in the literature: confounded by long-term (or average) market timing

1

10, 1

0

ts s

tsefwa t s

r rr

e dM M

B Be d

Table 8: Horse Race(OLS a la Baker and Wurgler, 2002, Y=book average)

EventYear t N

InitialMtBefwa, t-1

B&WMtBefwa, t-1 MtB t-1 PPE t-1 Profit t-1 LnSale t-1

Adj. R2

Year 3 5,904 -1.81 -3.08 0.12 -0.18 3.56 0.179

(-6.68) (-10.37) (10.00) (-9.13) (20.32)

-2.16 -2.92 0.12 -0.19 3.54 0.186

(-9.23) (-10.79) (9.58) (-9.77) (20.22)

-2.36 0.29 -3.01 0.12 -0.19 3.54 0.186

(-6.40) (0.68) (-10.06) (9.57) (-9.76) (20.22)

Year 5 4,799 -2.02 -3.06 0.12 -0.24 3.05 0.177

(-7.93) (-10.32) (8.70) (-9.25) (15.71)

-2.11 -3.43 0.11 -0.24 3.05 0.188

(-11.14) (-13.38) (8.12) (-9.18) (15.80)

-2.32 0.34 -3.57 0.11 -0.24 3.06 0.188

(-7.76) (0.85) (-11.80) (8.12) (-9.16) (15.84)

It is growth type and not market timing that best explains capital structure.

EventYear t N

InitialMtBefwa, t-1

B&WMtBefwa, t-1 MtB t-1 PPE t-1 Profit t-1 LnSale t-1

Adj. R2

Year 10 2,707 -2.67 -2.72 0.07 -0.39 2.14 0.170 (-7.66) (-6.17) (4.23) (-10.10) (8.80)

-2.11 -3.68 0.07 -0.37 2.24 0.173 (-9.35) (-9.29) (3.87) (-9.53) (9.42)-1.40 -1.27 -3.13 0.07 -0.38 2.16 0.175 (-4.08) (-2.39) (-6.76) (3.86) (-9.66) (8.93)

Year 15 1,490 -1.27 -4.54 0.04 -0.40 2.33 0.145 (-2.71) (-7.35) (1.52) (-7.61) (7.15)

-1.45 -4.82 0.03 -0.39 2.36 0.153 (-4.98) (-8.24) (1.28) (-7.30) (7.46)-1.46 0.03 -4.84 0.03 -0.39 2.37 0.153 (-3.80) (0.04) (-7.47) (1.27) (-7.28) (7.35)

FM 1,790 -1.79 -1.83 0.06 -0.51 2.79 71-05  (-19.27) (-3.78) (7.27) (-9.52) (10.73)

  -1.63 -2.02 0.05 -0.51 2.87   (-22.66) (-3.69) (6.45) (-9.74) (10.99)  -1.65 -0.07 -1.84 0.05 -0.51 2.87   (-9.84) (-0.30) (-3.56) (6.57) (-9.67) (11.41)

ConclusionA Growth Type Story for leverage persistence:Growth type gives rise to persistently distinct market imperfections across

firms and presets persistently distinct financing behavior in response

Findings: low (G1), mixed (G2) and high growth type (G3)1. Future leverage ratios persistently differ across the three Initial growth

types—explaining leverage persistence parsimoniously!

2. The growth types are rooted in firm fundamentals:(1) G1=tangible-growth type, low sales growth and cash, high and stable

profits, div payerG3=intangible-growth type, high sales growth and cash, high uncertain profits, div non-payer

(2) Equity issues and changes in retained earnings largely offset for G3 where R&D investments are relentless

Conclusion (Cont’d)3. Market Timing is preset by growth type.

During market price upturns, new equity issues↑ for all growth types, but the reasons differ across growth types:(1) AI of Type A ↓ good for low growth type (G1)(2) AI of Type B ↑ good for high growth type (G3)

consistent with the generalized Myers-Majluf viewA growth-type-based pecking order in financing

4. A two-horse race shows It is growth type and not market timing that best explains long-run capital structure. Fair but not exploitable view on market timing is compatible with our growth type determination story.

5. In short, if managers cannot change growth type, they cannot alter long-run capital structure through market timing.

Total Debt Short-term Debt (34) + Long-term Debt (9)Market Equity Stock Price (199) * Common Shares Outstanding (54)

Asset Total assets (6)

Leverage (L)(i)Book Leverage(ii)Market Leverage

Total Debt / AssetTotal Debt / (Total Debt + Market Equity)

Market-to-Book (MtB) [Market Equity + Total Debt + Preferred Stock (10) – Deferred Tax (35)] / Asset

Tangibility (Tang) [Inventory (3) + Property, Plant and Equipment (8)] / Asset

Profitability (Profit) Operating Income before Depreciation (13) / Asset

Firm Size (LnSize) Natural log of (Asset * 1,000,000), where Asset is deflated by GDP deflator (in 2000 dollar)

Ind_median Median industrial leverage according to the Fama and French classification of 38 industries

DivPayer Dummy variable: 1 for dividend payer and 0 for non-payer

Asset Growth Rate [Assett – Assett-1] / Assett-1

Sales Growth Rate [Salest (12) – Salest-1] / Salest-1

Investment Expenditure

Capext (128) / Assett-1; R&Dt (46) / Assett-1 (Note that only R&D missing values are replaced by zero.)

Cash Holdings Casht (1) / Assett-1

Propensity to Pay (PTP), or %Payers

The percentage of dividend payers (of a firm group)

Debt The change in Total Debt, or net debt issue

Net Equity (i)Book Value(ii)Market Value

The change in net equity, or net equity issue(i)[ of Common and Preference Stock (108)– Purchase of Common and Preference Stock (115)](i)[Sharest (25) * Adjustt (27) – Sharest-1 * Adjustt-1] * [Pricet-1 (199)/ Adjustt-1 + Pricet / Adjustt ]/2

RE The change in retained earnings (36)

Thank You!

Does competition from new equity mitigate bank rent-seeking at high growth firms? Insights from Japanese Data

Xueping Wu, Piet Sercu and Jun Yao Information asymmetry

mainly about growth Debt overhang likely Monitoring and information

production needed

Limited investment opportunities

Screening needed

Better prospects and reputation

Information asymmetry mainly about assets-in-place

Equity as last resort

Bank loans Total debt

Market-to-book ratio (Growth)

Banks’ rent-seeking attempts

Competition from bonds Competition from equity

E =1.00 E =2.00 E =3.00 E =1.00 E =2.00 E =3.00

sA=

0.50 1.41 0.73 0.23 0.96 0.64 0.330.75 1.26 0.54 0.07 0.92 0.58 0.261.00 1.06 0.29 -0.21 0.83 0.48 0.171.25 0.83 -0.08 -0.66 0.74 0.35 0.001.50 0.49 -0.59 -1.33 0.62 0.16 -0.221.75 0.15 -1.15 -2.23 0.49 -0.07 -0.452.00 -0.28 -1.93 -3.38 0.31 -0.35 -0.782.25 -0.73 -2.80 -4.57 0.15 -0.63 -1.19

-2.00 -7.24 -27.82 -45.25 -1.05 -5.65 -10.10-1.00 -1.37 -10.08 -19.63 0.60 -1.84 -4.850.00 1.22 -1.59 -4.60 1.42 0.27 -0.861.00 1.10 0.24 -0.19 0.93 0.41 0.122.00 0.16 0.05 0.00 0.17 0.06 0.023.00 0.00 0.00 0.00 0.01 0.01 0.004.00 0.00 0.00 0.00 0.00 0.00 0.005.00 0.00 0.00 0.00 0.00 0.00 0.00

0.50 -0.28 -1.17 -1.33 0.12 -0.20 -0.341.25 1.41 0.73 0.21 1.02 0.79 0.372.00 1.63 1.59 0.99 1.28 1.34 1.022.75 1.51 1.97 1.44 1.24 1.58 1.453.50 1.21 1.96 1.62 1.18 1.66 1.684.25 1.67 1.82 1.73 1.15 1.72 1.745.00 0.97 1.73 1.77 1.06 1.65 1.805.75 0.92 1.63 1.69 0.92 1.63 1.76

Panel A: Announcement Return and Relative Uncertainty about Growth

Panel C: Announcement Return and Expected Growth and Uncertainty in Tandem

Panel B: Announcement Return and Expected Growth

5A

B B s

10A

0.51 wE/10,c 1.00,σ A

0.51 wE/10,c 1.00,σ 1.00,B B

0.51 wE/10,c 1.00,σ 1.00, BA s B

Wu and Wang (JCF 2005)