Organizational Capital and Loan Financing

Adnan Anil Isin

Associate Research Fellow at Tax Administration Research Centre, University of Exeter

Examining syndicated loan facilities I show that investment in knowledge-based organizational capital,

the most important multiplier of intangible assets, is associated with lower loan spreads and covenant

intensity while enabling access to larger loan facilities. In particular, the price-based contractual benefits

associated with organizational capital are more pronounced for firms with larger institutional ownership

and analyst coverage. Moreover, these contractual benefits are not statistically different for firms that

operate in states that apply stricter non-compete employment agreements – legal structures argued to

shield migration of knowledge-based capital (if any) via key-talent outside options – compared to those

that do not. Accordingly, lenders seem to acknowledge systematic shift in investments towards

knowledge-based organizational capital and associated value creation thereof. Notably, however, rather

than focusing on state-level restrictions on employment mobility, they seem to concentrate more on

governance structures that help ensure credible investments in intellectual capital. Overall, the analysis

reveals a direct link between internally-generated intangible capital stock and loan financing that

extends beyond the arguments that historically linked asset tangibility with firm-level financial

flexibility.

Keywords: Organizational Capital, Intangible assets, Cost of bank debt, Agency costs, Asset

tangibility, Non-compete agreements, Contract design

Tax Administration Research Centre, University of Exeter Business School, Streatham Court, Rennes Drive,

Exeter, United Kingdom, EX4 4PU. E-mail: A.A.Isin@ex.ac.uk TEL: +44 (0) 1392 722128

1. INTRODUCTION

Corporate investment environment has been increasingly shifting towards knowledge-based “organizational

capital” as opposed to physical and fixed capital investments over the recent years (e.g., Atkeson and Kehoe,

2005; Hulten and Hao, 2008; Corrado, Hulten and Sichel, 2009; Corrado and Hulten, 2010; OECD, 2013).1

Research shows that, organizational capital, an accumulated know-how that efficiently connects tangible and

intangible resources in the production process (e.g., Evenson and Westphal,1995; Lev and Radhakrishnan, 2005;

Evans, Lev and Radhakrishnan, 2016), has become a crucial input in corporate value creation and sustainable

growth (e.g., Hulten and Hao, 2008; Corrado et al., 2009; Corrado and Hulten, 2010). On the flipside, given its

partly firm and partly talent-specific nature (Eisfeldt and Papanikolaou, 2013; 2014), investment in

organizational capital (OCAP) is shown to be relatively risky and fragile from shareholders perspective due to

the uncertainties related to the allocation of associated (related) cash flows in the presence of key-talent outside

options (e.g., Samila and Sorenson, 2011; Eisfeldt and Papanikolaou, 2013; 2014; Boguth, Newton and

Simurtin, 2016).

On the other hand, there is no direct evidence as to how creditors perceive organizational capital which

comprise around 30% of corporate-level intangible assets (Corrado et al., 2009). This is an important

shortcoming given creditors’ preference towards asset tangibility under the incomplete contracting theory (e.g.,

Shleifer and Vishny, 1992; Hart and Moore, 1994; Holmstrom and Tirole, 1997; Almeida and Campello, 2007;

Campello and Giambiona, 2010) and the systematic corporate-wide switch towards “hard-to-value” investments

in intellectual capital over time (e.g., Fu, Huang and Wang, 2015). Although recent evidence documents a

positive link between intangible capital and financial flexibility (Attig and Cleary, 2014; Lim, Macias and

Moeller, 2016) these studies focus on aggregate leverage levels instead of direct bank financing costs associated

with organizational capital. Similarly, while investigating direct costs of bank capital, Loumioti (2012) has a

particular focus on the collateral value of intangible assets and do not investigate organizational capital stock, a

subset of intangible capital stock. In fact, both Lim et al (2016) and Loumioti (2012) concentrate on intellectual

capital generated as part of business combinations and/or acquisitions. Accordingly, pertinent to the

understanding of the contracting theory, creditors’ perception of investment in internally-generated intellectual

capital remains as an empirically relevant and timely question. This paper aims to answer this particular

1 For the purposes of this paper I will be using terms organizational capital, knowledge-based and intellectual

capital interchangeably. Also I will use abbreviation, OCAP, to refer to organizational capital.

question and establish a direct link between organizational capital and contractual (price and non-price) terms

associated with bank financing.

Using 8944 syndicate loan facilities from 1994 to 2015 for non-financial U.S. public firms, I document an

economically meaningful negative link between investment in organizational capital and cost of bank financing.

Specifically, a standard deviation increase in organizational capital stock helps reduce loan spreads by 4.70BPS

on average. For a typical loan facility in the sample ($600MN) this corresponds to $1.14MN savings in interest

costs over the average loan period (4 years).

Next I examine whether the contractual benefits associated with organizational capital are more pronounced for

firms with stronger external monitoring mechanisms intact. In doing so, I focus on institutional ownership and

the number of analysts following. The literature contends that these governance mechanisms mitigates

opportunistic earnings management (e.g., Chung, Firth and Kim, 2002; Chen, Harford and Lin, 2015), excessive

managerial compensation and rent extraction (e.g., Hartzell and Starks, 2003; Chen, Harford and Lin, 2015)

managerial short-termism (Wahal and McConnell, 2000). The literature also documents that these governance

mechanisms facilitate enhanced information processing and stock market efficiency (e.g., Brennan and

Subrahmanyam, 1995; Jiambalvo, Rajgopal and Venkatachalam, 2002; Piotroski and Roulstone, 2004; Velury

and Jenkins, 2006; Ellul and Panayides, 2009), managerial incentive alignment (e.g. Hartzell and Starks, 2003),

and budgetary discipline (e.g., Harford, Mansi, and Maxwell, 2005; Dittmar and Mahrt-Smith, 2007) as well as

financing and investment flexibility (Derrien and Keckés, 2013). In fact, more recent work documents a direct

and a negative link between institutional ownership with costs of loan financing (e.g., Elyasiani, Jia and Mao,

2010; Roberts and Yuan, 2010). In line with the above empirical evidence I find that the negative link between

organizational capital and bank financing is more pronounced for firms with stronger governance mechanisms.

Moreover, I examine, from lenders’ perspective, whether and if so to what extend they perceive investment in

knowledge-based capital as firm-specific and sustainable or talent-specific and subject to competitive

penetration via employment exchange. This question is particularly relevant given that the literature is split into

two opinions where group of authors argue that investment in OCAP is firm-specific and remain so regardless of

the key-talent outside options (e.g., Atkeson and Kehoe, 2005; Lev et al., 2009; Evans et al. 2016) while others

argue that its partly firm and partly talent-specific (Samila and Sorenson, 2011; Eisfeldt and Papanikolaou,

2013; 2014; Boguth et al., 2016). Thus, while the former statement argues that ownership of cash flows

generated from organizational capital is firm-specific and resilient to competitive threats, the latter argues

otherwise. Moreover, the answer to the above question is of particular importance from lenders’ point of view as

they hold asymmetric claims on firm performance which limits their participation on upside potential while

facing substantial exposures to downside risks. If knowledge-based intellectual capital is concentrated on “the

highly-skilled few” that holds flexible outside options, then returns on these investments are likely to face

competitive risks.

Given the above discussion I take on a conservative approach and perceive that the organizational capital is

partly firm and partly talent-specific in nature (e.g., Samila and Sorenson, 2011; Eisfeldt and Papanikolaou,

2013; 2014; Boguth et al., 2016). Accordingly, I use state-level intensity of non-compete agreements (NCAs) as

an exogenous input (e.g., Gillan, Hartzell and Parrino, 2005; Garmaise, 2009) that controls for (state-level)

restrictions on employee-mobility. These legal frictions on employee outside options has received increasing

public (e.g., Jasper, 2010; Muro, 2016; Viswanatha, 2016; Ben-Shahar, 2016), regulatory (DoT, 2016;

Whitehouse, 2016) and academic (e.g., Garmaise, 2009, Starr, 2016; Starr, Ganco and Campbell, 2016; Younge

and Marx, 2015; Klasa et al., 2015) attention over the recent years. In essence, non-compete covenants are legal

employment restrictions with the ultimate aim to avoid/delay migration of intellectual capital, including but not

limited to trade secrets, specialized know-how, commercial relationships and/or contacts with specific

prospective or existing customers, to competing firms through employment exchange. Given that much of the

innovation is derived from tacit knowledge that is difficult to patent-protect (e.g., Samila and Sorenson, 2011),

these employment frictions can safeguard firms against knowledge-based capital leakage which may occur via

key-talent transfers to competitors. By efficient protection of cash flows generated from knowledge-based

capital, these non-compete clauses thus may help enhance returns on firm-specific investment in intellectual

capital. For example, research shows that NCAs limit within-industry job transfers and increase employment-

tenure particularly among the key-skilled human capital (e.g., Garmaise, 2009; Starr, Prescott and Bishara,

2016). Thus, controlling for state-level restrictions on employee-mobility allows me to examine creditors’

perspective on risks associated with cash flow ownership from investment in OCAP.

In doing so, I develop an initial NCA index (NCEI) in the light of the guidance provided in Malsberger (2013).

Next, I track, to the best of my capacity, all the legislation changes in state-level non-compete agreements

(NCAs), manually read each “passed” bill when necessary and update NCEI index for the years 2013 to 2015.

In additional analysis I also utilize non-compete index provided by Garmaise (2009) by limiting the sample for

periods between 1994 and 2004. To control for self-selection into high-NCA-intensity states by firms with large

investments in OCAP, I match firms based on their observable characteristics, leaving state-level ranking on

NCEI as a treatment effect. The analysis using this matched sample indicates that the negative link between

organizational capital and bank financing is neither statistically nor economically more pronounced for states

with stricter non-compete agreements. These results hold under models that control for multi/single state firms,

external corporate governance measures, customer-supplier links and alternative empirical settings that examine

the effects of NCA intensity.2

Finally, I investigate the link between non-price loan terms and organizational capital. Following the past

literature, I consider greater number of restrictive covenants (e.g., Demiroglu and James, 2010) and the

syndicate-lead ownership ratio (e.g., Sufi, 2007; Mora, 2015) as contractual terms that reflects borrower

riskiness. Similarly, I also control for the total number of bank/lender participation at lead-level (e.g., Isin,

2016). If lenders perceive firms with larger investments in intellectual capital riskier than those with more

physical capital investments, syndicate arrangers may place greater number of restrictive covenants, hold larger

portion of the loans provided and form loan facilities with larger number of lead arrangers. And finally, I control

for the average loan size. One would expect to see a negative/(positive) relation between heavier reliance on

intangible-based investments and average loan size to the extent that lenders consider these investments

riskier/(safe) investments in relative terms. In that sense the analysis aims to reveal, if any, strictness in non-

price loan terms that lenders might trade off with price-based terms to control for potential risks inherent in

investments in intellectual capital (i.e., employee outside options). Results show that, on average, loans issued to

firms with larger investment in knowledge-based organizational capital contain lower number of restrictive

covenants, particularly for firms that operate in states with stricter NCA rules.3 Moreover, on average, sizeable

investment in organizational capital is also associated with larger funding arrangements regardless of state-level

NCA intensity. Notably, I fail to find supporting evidence where asset tangibly is associated favorable non-price

loan terms (e.g., looser covenant clauses and/or access to larger financing) but some evidence documenting

potential risks associated with R&D intensive firms.

This paper makes several noteworthy contributions to corporate finance, banking and labor economics

literatures. First, the paper extends recent evidence that presents a positive link between intangible assets and

firm-level access to external capital (Loumioti, 2012; Lim et al., 2016). Specifically, the analysis presents a

2 In un-tabulated analysis I also use the non-compete index provided by Garmaise (2009) by limiting the sample

into years including 2004 and before, my own non-compete index by limiting the sample into years following

2001. My conclusions do not change under these alternative sample and modelling considerations. 3 This difference is statistically significant at p<0.0001

direct negative link between organizational capital, as the most important multiplier of total intangible assets

(Corrado et al., 2009), and price and non-price terms associated with loan financing.4 Therefore, the analysis

shows that the economic competencies achieved via knowledge-based capital investments extend beyond the

collateral value of intangible assets, particularly for better governed firms – complementing the evidence

Loumioti (2012). Moreover, having access to larger loan facilities with less strict covenant terms indicates that

investment in knowledge-based capital is no more risky than physical, fixed capital and R&D investments.

Overall, despite being kept off the books, lenders seem to acknowledge organizational capital as an important

economic competency that generates sustainable competitive advantage and a stable revenue stream particularly

for firms with robust governance structures that help ensure credible investments in intellectual capital. The

evidence, therefore, is timely and relevant in the light of systematical increase in intangible capital and ongoing

debate as to whether and to what extend these investments should be capitalized. Specifically, although results

confirm the general argument that lenders rationally assess value-generation from organizational capital as a

form of intangible capital (Skinner, 2008), I fail to document such efficiencies for internally-generated

intangible capital as a result of R&D expenditures/capital stock.

Finally, the analysis also intersects empirical loan contracting research with labor economics. Specifically,

results show that lenders, unlike to shareholders (Eisfeldt and Papanikolaou, 2013; 2014), do not seem to take

on board potential cash-flow-ownership risks, that might stem from skill-transfer via key-talent outside options,

associated with knowledge-based capital. A combination of alternative factors might drive these observations.

First, creditors might not be paying enough attention to potential cash-flow-ownership risks than they actually

should or that they perceive organizational capital as firm-specific regardless of the key-talent outside options

(e.g., Atkeson and Kehoe, 2005; Lev et al., 2009; Evans et al. 2016). And finally, non-competes only delay skill-

transfer and might be un-enforceable if these covenants aim to restrict competition in an unreasonably large

territory and/or duration which may prove these employment restrictions ineffective (eventually) in the long run,

at least from the lenders’ point of view. Hence, in the context of labor economics, the evidence provided in this

4 Although Loumioti (2011) focuses on direct costs associated with bank financing, she finds larger spreads

associated with loans collateralized intangible assets which renders support for the argument that intangible are

riskier, on average, despite supporting additional external financing.

study renders some credit to the recent arguments challenging the effectiveness of non-compete agreements

(e.g., DoT 2016; Whitehouse, 2016).5

2. INSTITUTIONAL KNOWLEDGE and HYPOTHESIS DEVELOPMENT

The past three decades have witnessed intangible forms of investments overtake tangible/fixed capital

investments in magnitude and as the main driver of value creation both domestically (U.S.) (Corrado et al.,

2009; Corrado and Hulten, 2010; Fu, Huang and Wang, 2015) and internationally (Corrado et al., 2013; Corrado

et al., 2014). Particularly with the help of globalization U.S. firms have increasingly outsourced their capital-

intensive production functions to developing countries and focus on knowledge-intensive capital that oversees

product design, development and distribution. An everyday example would be to look behind any Apple product

(i.e., mobile phones) which clearly states geographical separation of the knowledge-intensive product design

and development functions, which generates the core technological competency of the end-product, and the

capital-intensive manufacturing function.

Accordingly, it is somewhat surprising that given the role intangible capital plays in value creation, the large

majority of the in-house generated intangible assets are not reflected in firms’ balance sheets. Some argue that

not capitalizing intangible assets on the balance sheet hampered the relevance and the usefulness of accounting

information (Lev and Zarowin, 1999) given that intangible assets explain around 45% of the total market values

of 617 R&D intensive firms (Hulten and Hao, 2008). Nonetheless, under both US GAAP and IFRS intangible

assets are defined as “non-monetary” assets with no physical substance (E&Y, 2011). There are several

alternative factors cited for non-capitalization of intangible investments. For example, FASB’s SFAS 2 argue on

the lack of casual link between R&D expenditures and measurable future benefits in the forms of increased

sales, earnings and market shares. Kanodia et al. (2004) outlines i) impairment of the reliability of reported

accounting numbers due to errors associated with measuring intangibles and ii) ) expansion of opportunities for

earnings management using intangible assets as relevant reasons why FASB is reluctant to capitalize intangible

investments. Moreover, the authors argue that the recognition of intangibles should be conditional on the weight

such assets receive in a firm’s capital stock under the current “expensing regime”. Skinner (2008) argues that

capital markets already do an efficient job of financing firms with intangible capital without the need of a

further accounting regulatory overhaul. Moreover, investment in intangible capital is assumed to be risky in

5 Alternatively, aggregate index of non-compete intensity might not fully capture the effects of such clauses on

loan contracts (e.g., Starr, 2016) particularly for a syndicate loan sample that is notoriously biased towards

larger corporations. Hence, the analysis regarding non-competes may warrant further investigation.

nature with limited certainty as to the success of the project(s) undertaken. For example, Kothari et al. (2002)

demonstrate that R&D investments possess much higher uncertainty regarding future accessible benefits in

comparison to fixed capital investments. Similarly, Hall and Lerner (2010) argue that innovation is difficult to

finance given intense reliability on human capital and/or other non-physical assets and informational frictions

thereof.

Traditionally, the literature has overwhelmingly focused on asset tangibility arguing that under incomplete

contract theory and limited enforceability firm-level access to external financing is not frictionless (Hart and

Moore, 1994; Holmstrom and Tirole, 1997). Under this framework, tangible assets (which are verifiable by

courts) are likely to have greater collateral value in comparison to intangible assets under low states of the world

(e.g., Kanodia et al., 2004). Therefore, firm-level access to external capital is show to be more flexible when the

financing can be backed with easier-to-value and deploy assets rather than hard-to-value and verify intangibles

(e.g., Almeida and Campello, 2007; Benmelech, 2009; Benmelech and Bergman, 2009; Campello and

Giambona, 2010).

Research on intangible investments and external financing, on the other hand, is at its preliminary stage given

researchers have overwhelmingly focused on asset tangibility. To the best of my knowledge there are two

concurrent working papers examine intangible assets and capital structure and bank financing. Among these,

Lim et al (2015) show that on a per dollar basis across all firms intangible assets support around 3/4th as much

debt financing as tangible assets do. They also show that banks perceive firms with relatively few tangible assets

as riskier borrower. Accordingly, loans provided to firms with larger reliance on intangible assets have shorter

maturities, and are mostly secured bank debt with likely convertible features. On the other hand, Loumioti

(2012) examines collateral value of intangible assets and finds that around 21% of the total loans originated

between 1996 and 2005 use intangible assets as part of the collateral pool. Moreover, loans that use intangible

assets as part of a collateralization system increase the total outstanding loan size by approximately 18% but

also the average loan costs by 74BPS ($4.1MN in additional interest expenses).

Overall, the intake from these two studies is that intangible assets have operational (i.e. cash flow generation)

and collateral value that collectively help increase firm-level borrowing capacity. In doing so, however, both

studies document riskier profiles for firms with greater reliance on intangible assets from lenders’ point of view.

Given this evidence one can reasonably expect to observe larger loan spreads for loans originated for firms with

larger investments in knowledge-based organizational capital. This line of thinking would also be logically

aligned with the arguments made from equity investors’ point of view.

On the other hand, both papers use intangibles that are capitalized, predominantly as part of business

combinations and/or acquisitions, and do not generally consider intangible capital that is internally generated

and expensed due to accounting restrictions. In essence, intangible assets have fine-line differences both in their

capacity and accounting recognition. For example Corrado et al. (2005) segregate knowledge-based capital into

i) computerized information, ii) innovative property and iii) economic competencies. Organizational capital lies

within economic competencies heading and represents firm-level capacity to effectively utilize all the functions,

systems and information processed in the former two headings including research and development of new

products, designs, software and databases. In fact, Lev and Daum (2004) argue that the true value creation via

intangible investments occur when they are used in combination with other production factors.

Moreover, investment in knowledge-based capital has been the primary engine for economic growth over the

past thirty to fifty years (e.g., Corrado et al., 2009; Corrado and Hulten, 2010; Fu, Huang and Wang, 2015) and

now reflects around 45% of the total market values of 617 R&D intensive firms (Hulten and Hao, 2008). Lev et

al. (2016) perceive organizational capital as knowledge, systems and procedures that effectively connects

tangible and intangible resources in the production process. This line of thinking, on the other hand, indicates

that lenders, on average, should benefit from investment in organizational capital given that economic

competencies obtained thereof extends into other forms of intangible assets/investments. Moreover, given this

systematic shift in firm-level focus on knowledge-based capital as the primary engine for value creation I expect

lenders to adjust their credit rationing accordingly and factor in the “new world” investment order in their

lending practices and contractual design choices. In the light of the above discussed two competing arguments I

conjecture that;

H1: On average, investment in organizational capital is not riskier than other forms of tangible/fixed capital

investments and do not warrant higher loan spreads.

An alternative measure where lenders can control for additional risks they assume, if any, for firms with greater

reliance on intangible capital would be to use stricter non-price terms (e.g., Nini, Smith and Sufi, 2009;

Demiroglu and James). We can observe stricter non-price loan terms regardless of loan spreads as price and

non-price loan terms might, under certain market conditions, move in opposite tandem. For example, recent

evidence shows that lenders, particularly under low-yield credit conditions, might switch into strategies where

they trade-off some of the non-price terms for better yields (Stein, 2013; Becker and Ivashina, 2016).

Accordingly, for testing the baseline analysis examining the link between OCAP and non-price loan terms I

make no directional predictions.

H2: On average, investment in organizational capital is not associated with stricter non-price loan terms in

comparison to other forms of tangible/fixed capital investments.

In their seminal paper Jensen and Meckling (1976) argues that separation of ownership and control creates

informational and incentive-related conflicts of interests among management, shareholders and creditors.

Accordingly, managements could take on riskier and self-interested ventures that would deviate from value

maximization at the expense of shareholders, minority investors and/or bondholders. On the other hand, Jensen

and Meckling (1976) also argue that agency costs associated with the separation ownership and control would

lead to monitoring activities specialized to those institutions and individuals who possess comparative

advantages in these tasks (see page 354). Two of such monitoring groups emerged are financial analysts and

institutional investors. For example, evidence shows that institutional ownership mitigates opportunistic

earnings management (e.g., Chung, Firth and Kim, 2002), excessive compensation (e.g., Hartzell and Starks,

2003) and managerial short-termism (Wahal and McConnell, 2000) and facilitates enhanced information

processing (e.g., Jiambalvo, Rajgopal and Venkatachalam, 2002; Piotroski and Roulstone, 2004; Velury and

Jenkins, 2006 ), incentive alignment (e.g. Hartzell and Starks, 2003), managerial/budgetary discipline (e.g., .

Harford, Mansi, and Maxwell, 2005; Dittmar and Mahrt-Smith, 2007). More recent work also illustrates

institutional ownership with lower costs of bank financing (e.g., Elyasiani, Jia and Mao, 2010; Roberts and

Yuan, 2010). Similarly, greater analysts coverage lowers opportunistic earnings management and rent extraction

(e.g., Yu, 2008; Chen, Harford and Lin, 2015) increases information processing and stock market efficiency

(e.g., Brennan and Subrahmanyam, 1995; Ellul and Panayides, 2009) and facilitates higher corporate investment

and financing (Derrien and Keckés, 2013). In the light of this evidence, I anticipate creditors to acknowledge

monitoring efforts by institutional investors and/or financial analysts and expect that

H3: On average, regardless of the outcome in H1, price-based loan contracting terms are more favorable for

firms with stronger external governance proxies intact.

H4: On average, regardless of the outcome in H2, investment in organizational capital is associated with looser

non-price loan terms for firms with stronger external governance proxies intact.

The final factor to consider is the observation that there is no clear agreement as to whether and to what extent

the investment in knowledge-based organizational capital is talent and/or firm-specific (Monzur, 2015). One line

of thinking argues that these investments are firm-specific and disregards the key-talent outside option (Atkeson

& Kehoe, 2005; Lev et al., 2009; Lev et al. 2016). For example, Lev et al (2016) argue that organizational

capital is firm-specific and cannot be easily imitated by competitors. On the other hand, an alternative line of

thinking argues that at least a portion of intellectual capital is linked to key-talent (Samila and Sorensen, 2009;

Eisfeldt and Papanikolaou, 2013; 2014; Boguth et al., 2016). In fact, Boguth et al. (2016) show that sudden

departures of CEOs have significantly larger/(smaller) disruption costs to organizational capital for firms with

higher/(lower) organizational capital stocks. This evidence renders credit to the argument that investors require

additional risk premiums for bearing this form of cash flow ownership risk (Eisfeldt and Papanikolaou, 2013).

Given the above inconclusive evidence I side with the argument that knowledge-based capital is partly firm and

partly talent-specific but these components might differ cross-sectionally.

One potentially effective measure that can help shield migration of knowledge-based capital (if any) via key-

talent outside options is the state-level non-compete agreements which place employment restrictions on certain

employees from switching to competitors. Accordingly, I empirically test, from lenders’ perspective, what

portion of the investments in knowledge-based organizational capital matters the most. In doing so I utilize

NCEI index which estimates the intensity of state-level non-compete agreements over time which allows me to

introduce an exogenous factor to test the acclaimed effects of potential key-talent outside options. These laws

govern state-level frictions with the ultimate aim to avoid/delay migration of intellectual capital, including but

not limited to trade secrets, specialized and unique know-how, commercial relationships or contacts with

specific prospective or existing customers to competing firms via employment exchange. Although on macro

scale these clauses received some fair criticism (Gilson, 1999; Amir and Lober, 2013; Amir and Lober, 2016;

DoT, 2016; Whitehouse, 2016), on firm-level analysis recent evidence documents value generation through

wage-specific cost reductions (Kim and Marschke, 2005; Starr, 2016) and increased investment in human

capital (e.g., Starr, 2016).6

Accordingly, if lenders believe that an important portion of organizational capital is talent-specific they might

perceive firms with greater reliance on knowledge-based organizational capital to be more sensitive to

competitive threats that hamper accumulated investment in intellectual capital as a results of key-talent

6 I discuss different aspects of state-level non-compete agreements in section 3.2.

departures. If so they may consider state-level NCAs as an important mechanism that mitigates competitive

risks firms with heavy investments in intellectual capital face. In that case loans might favorably reflect state-

level NCA rules that safeguard intellectual capital investments from competitive threats stemming from

employment exchange. If on the other hand, lenders hold the view that investment in knowledge-based capital is

largely firm-specific with little or no concern of competitive threats associated with key-talent outside option I

expect to see no difference in loan spreads for firms that otherwise have similar interests in knowledge-based

capital. Accordingly, my seventh and eight hypotheses argue that;

H5: On average, lenders perceive investments in organizational capital as less/(more) risky for firms that

operate in states that apply stricter/(looser) state-level non-compete agreements.

H6: On average, investment in organizational capital is associated with looser/(stricter) non-price loan terms for

firms that operate in states with stricter/(looser) NCA rules.

3. RESEARCH DESING and SAMPLE SELECTION

3.1. Measuring Organizational Capital

In measuring organizational capital I follow Eisfeldt and Papanikolaou (2013) and estimate the stock of

organizational capital accumulated over time using the following model that use Selling, General and

Administrative (Compustat: xsga) to calculate the organizational capital and offers a proxy level of accumulated

organizational capital should such investments would have been capitalized.

𝑂𝐶𝐴𝑃𝑖,𝑡 = (1 − 𝑑𝑂𝐶 ) × 𝑂𝐶𝑖,𝑡−1 +𝑆𝐺&𝐴𝑖,𝑡

𝑐𝑝𝑖𝑡 (1)

𝑂𝐶𝐴𝑃𝑖,0 =𝑆𝐺&𝐴𝑖,1

𝑔+𝑑𝑂𝐶 (2)

In the above model OCAP is the organizational capital for firm i at time t. The model recursively estimates firm-

level organizational capital stock by starting from year 0 initial estimate of organizational capital eq. (2). In

estimating this initial organizational capital stock we assume 10% growth rate in investment in organizational

capital and 15% depreciation rate. Firm-year level organizational capital is then estimated by accumulating

inflation-adjusted (cpi) SG&A and appropriately depreciating these investments over time. In the above

calculation all the missing values of SG&A is treated as zero as in Eisfeldt and Papanikolaou (2013). This

estimate of organizational capital is then scaled by book value of assets and termed OC to reflect firm-level

organizational capital stock. In the empirical models I also use industry adjusted organizational capital IAOC

where I de-mean firm-level organizational capital (OC) with the industry-year mean levels based on 2-digit SIC

codes. In additional analysis I also use dummy indicator (OCQ) that takes the value of 1/(0) if firm-level

organizational capital is above/(below) sample median levels.

3.2. Measuring Non-Competition Enforceability Index

The starting point of the NCEI index is the analysis provided by Malsberger (2013) who examines jurisdictional

variation in enforcement intensity of non-compete laws. The author details state-level applications of non-

compete agreements by providing insightful answers into twelve questionnaire from 2001 to 2013.7

Accordingly, the index can range from 0, for states that restricts non-compete agreements and a possible 12 for

states that has the strictest applications of such clauses. Following the years 2013 I track, to the best of my

capacity, all the legislation changes in state-level non-compete agreements (NCAs), manually read each

“passed” bill when necessary and update NCEI to cover years 2014 and 2015.8 There exists substantial amount

of heterogeneity among states as to the intensity of the NCA rules. For example. NCAs are banned in California

since the 19th century. In fact, the state bans three main types of NCA restrictions that covers issues where i) an

employee working for a competitor ii) an employee soliciting customer(s) and iii) an employee soliciting

another employee. Similarly, certain states have stricter applications of NCAs on certain income/occupation

groups while looser on the others. For example, Hawaii banned NCAs (H.B. No. 1090) restricting high-tech

workers from switching to competing firms in July 2015. On the other hand, Connecticut enacted legislation

(Public Act No. 16-95) that limits enforceability of NCAs on physicians for all contracts entered or extended on

or after July 2016.

Moreover, while some states had small to modest changes in NCA rules (e.g., Connecticut (2013 – Bill No.

693), Idaho (2008 – Senate Bill 1393) some have seen more radical changes. (e.g., Georgia – H.B. No. 30). For

example, new legislation in Idaho now clearly states that employers may use written agreement to protect their

legitimate business interests and if the restrictions are overbroad as to the covenants definition can be altered by

courts under “Blue Pencil” doctrine. On the other hand, Georgia has made it significantly easier for employers

to enforce restrictive covenants against former employees which reversed decades long of Georgia court

decisions in favor of the employees. The new legislation includes “Blue Pencil” doctrines, clauses that place

7 Note that Malsberger (2013) examines additional questions within the main questions and additional main

questions, such as applicability choice of law provisions, on top of those provided in Malsberger (2004). To

assure comparability and consistency with the index provided in Garmaise (2009) I stick to the original twelve

questions examined. 8 Interested readers can refer to Kenneth J. Vanko’s insightful blog http://www.non-competes.com/ for much

detailed guidance and discussions on the topic.

greater weight on the interests of the firms relative to those of employees and those that leave the burden of

providing reasonableness/unreasonableness of the covenant on the employee. Moreover, new rules also gives

less account on the economic hardship faced by the employee and accept continued employment to provide

sufficient consideration to support a covenant not to compete entered into after the employment relationship has

begun .

On the other hand, there are numerous cases where states have failed to pass bills that targets NCAs. Some of

these states include; Illinois, Michigan, Massachusetts and Wisconsin. These proposed changes in NCA rules

are excluded from the NCEI index estimations. On the other hand, some more recent state-level changes cannot

be applied to this paper given that the new legislation will be effective in 2016 or late 2015 (e.g., Alabama,

Arkansas; Oregon). I eliminate these changes from the NCEI index estimations given that they cover time

periods that are either beyond the range of this paper and/or impractical to apply.9 Finally, I do not adjust NCA

rule changes that covers non-corporate workers such as nurses and physicians (e.g., Connecticut, 2013).

In the main analysis I use this index to cover the total sample of observations from 1994 to 2015. However, in

addition to using my own NCA index I also utilize the index provided by Garmaise (2009) by limiting the

sample to years 1994 to 2004 as a layer of robustness analysis. I also use a hybrid versions (un-tabulated) where

in one I limit the sample years after 2001 to achieve consistency with Malsberger (2013) coverage period and in

the other I use the index provided by Garmaise (2009) until 2004 and switch to my own index following

(exclusive) the year 2005.

3.3. Empirical Modelling

I use the following primary model to test the effect of organizational capital on loan spreads.

𝐿𝑁𝑆𝑃𝑅𝐸𝐴𝐷𝑖,𝑡 = 𝑂𝐶/𝐼𝐴𝑂𝐶𝑖,𝑡−1 + 𝑇𝐴𝑁𝐺𝑖,𝑡−1 + 𝐶𝐴𝑃𝑋𝑖,𝑡−1 + 𝑅𝑁𝐷𝑖,𝑡−1 + 𝑃𝑃𝑃𝑖,𝑡 + 𝑁𝐿𝐷𝑖,𝑡 + (3)

+𝑁𝑃𝐴𝑅𝑇𝐷𝑖,𝑡 + 𝐿𝑃𝐶𝑇𝑖,𝑡 + 𝐿𝑅𝐸𝑃𝑖,𝑡 + 𝑁𝐶𝑂𝑉𝑖,𝑡 + 𝑇𝐸𝑅𝑀𝑖,𝑡 + 𝑅𝐸𝑉𝐷𝑖,𝑡 + 𝑆𝐸𝐶𝑈𝑅𝑖,𝑡

+𝐿𝑁𝐿𝑂𝐴𝑁𝑖,𝑡 + 𝑆𝐷𝐶𝐹𝑂𝑖,𝑡 + 𝐿𝑁𝑇𝐴𝑖,𝑡 + 𝑃𝑇𝑅𝑂𝐴𝑖,𝑡 + 𝐿𝑉𝑅𝐺𝑖,𝑡 + 𝑃𝐼𝐹𝑂𝑡 + 𝐴𝑄𝑖,𝑡

+𝐶𝐼𝑆/𝐶𝐹𝑁𝐴𝐼𝑡 + 𝑀𝑇𝐵𝑖,𝑡 + 𝑂𝑊𝑁𝑖,𝑡 + 𝐴𝐹𝑂𝐿𝑖,𝑡 + 𝑍𝑆𝐶𝑂𝑅𝐸𝑖,𝑡 + 𝑇 + 𝐼 + 𝜀𝑖,𝑡

9 For example, Arkansas Act 921 took effect on August 6, 2015. This would require updating the index for

Arkansas for the last five months of the year 2015. Such an effort, although a simple data management issue,

would not make too much sense given that the data coverage would be scarce and might include clauses that

lenders might not rapidly apply within their lending standards.

In the above model LNSPREAD is the natural logarithm of the loan spreads. OC and IAOC are one-year lagged

firm-level and industry-adjusted organizational capital stock scaled by total assets, respectively. TANG is the

asset tangibility and measured using net property, plant and equipment that is adjusted for depreciation

deduction (Compustat: PPENT) scaled by total assets. To the extent that asset tangibility mitigates borrower

risks via its collateral value I expect to observe a negative link between asset tangibility and loan spreads. CAPX

and RND are lagged capital and research & development expenditures that are both scaled by book value of

assets. PPP is the dummy indicator for loans that include performance pricing provisions. NCOV is the number

of covenants included in a loan. Past research documents that performance pricing provisions help alleviate

adverse selection and moral hazard problems between borrowers and lenders (e.g., Asquit, Beatty and Weber,

2005). Therefore, I expect to see, on average, a negative link between performance pricing provisions and loan

spreads. Covenant based terms can have positive and/or negative link with loan spreads given that the two might

move either in the same or alternative directions given macro and micro lending conditions (Stein, 2013).

I expect both covenant based non-price terms and PPPs to alleviate adverse selection and moral hazard problems

associated with bank financing and observe negative coefficients for these variables. NLD is the dummy

indicator which takes the value 1/(0) if the loan facility has greater/(less) than median number of lead arrangers

and controls for syndicate-lead-level risk diversification (Isin, 2016). NPARTD is a dummy indicator 1 if the

total number of participants in a syndicate loan is greater than the sample median, and zero otherwise. On

average, both lead (NLD) and non-lead (NPARTD) level ownership formation help alleviate information

asymmetries among the syndicate participants, hence I expect to find negative coefficients for these variables. I

also control for the proportion of loan held by the syndicate arrangers (LPCT) as in Sufi (2007) with the

exception that rather than averaging out the total LPCT among the syndicate-lead I use the total lead-level loan

ownership ratio. That is, if four lead arrangers hold half of the total loan amount the LPCT ratio corresponds to

50% rather that 12.5% (50%/4) for each lead arranger. I expect to see negative coefficients for the total lead-

level ownership ratio. In line with the past research (e.g., Denis and Mullineaux, 2000, Sufi, 2007) I also control

for lead arranger reputation where I classify top five syndicate arrangers in a given year in Thomson Deals

database as the most reputable lenders. I use a dummy indicator (LREP) that takes the value of 1 if the number

of reputable lenders in a given issue is above the sample-median and zero otherwise. I expect to see negative

coefficients for this variable.

I control for lead arranger reputation as an additional variable that mitigates agency conflicts among the lending

group as in the past research (Denis and Mullineaux, 2000, Sufi, 2007; Ball et al., 2008). Specifically, I classify

top five syndicate arrangers per given year in Thomson Deals database as the most reputable lenders. Next, I

identify loans with the number of reputable lenders in the top quartile of the total sample distribution (LREP). I

expect to see negative coefficients for this variable and average loan spread spreads. Finally, in line with the

past research (e.g., Sufi, 2007) I control for average loan maturity (TERM) and average loan size as the natural

logarithm of the loan amount (LNLOAN). I control for the difference between commitment loans and term

loans (e.g., Berger and Udell, 1995) by adding a dummy indicator (REVD) that takes the value of 1 if the loan is

a revolving credit facility and 0 otherwise. Finally, I also control for whether the loan is secured via collateral

(SECUR) as loan level control variables (e.g., Sufi, 2007).

Next, I include commonly-used firm level control variables. I control for firms size (LNTA) as the natural

logarithm of total assets (Compustat: at) and financial leverage (LVRG) as the total long term debt outstanding

(Compustat: dltt) divided by total assets. I control for firm-level profitability using total and foreign pre-tax

returns on assets (PTROA, PIFO) calculated as pre-tax income (Compustat: pi) and foreign pre-tax income

(Compustat: pifo) divided by total assets, respectively. I control for the accrual quality (AQ) as calculated in

Francis et al (2005) as a proxy for earnings quality and informational asymmetries. I expect to see a positive link

between low quality earnings (larger AQ) and loan spreads (e.g., Cook et al., 2015). I control for institutional

ownership (OWN) and the number of analysts following (AFOL) as measures of external corporate governance.

I expect stronger external governance to alleviate borrower-lender frictions and hence observe negative

coefficients for these variables (e.g., Roberts and Yuan, 2010). I control for the availability of firm-level growth

opportunities using market-to-book ratio (MTB). In addition I control for the effects of credit market tightness

using commercial and industrial spread (C&I spread) over federal fund rates (CIS) as in Harford et al. (2014).

Moreover, I control for the effects of the overall economic activity using Chicago Fed National Activity Index

(CFNAI). I expect to see both macro-level financial constraints to be priced in as additional loan spreads.

In order to test for non-loan terms associated with organizational capital I use an augmented version of the

model in equation (3)

𝑁𝑃𝑇𝐸𝑅𝑀𝑖,𝑡 = 𝑂𝐶/𝐼𝐴𝑂𝐶𝑖,𝑡−1 + 𝑇𝐴𝑁𝐺𝑖,𝑡−1 + 𝐶𝐴𝑃𝑋𝑖,𝑡−1 + 𝑅𝑁𝐷𝑖,𝑡−1 + ∑ 𝑋𝑖,𝑡 + 𝑇 + 𝐼 + 𝜀𝑖,𝑡 (4)

In the above model, NPTERM represent non-price loan terms that controls for covenant intensity (NCOV) and

loan size (LOANSIZE). X represents all the remaining control variables as defined in equation (3).

3.4. Sample Selection

The data for syndicated loan financing is obtained from Thomson One Deals database. This database has the

same coverage as SDC Platinum database that is widely used in empirical research. I match the loan data

obtained from this database with the financial data in Compustat database for the periods covering 1994 to 2015.

In forming the database eliminate financial and utility firms (SIC: 6000-6999, SIC: 4900-4999). This selection

criteria yields a maximum of 8944 loan-year observations and varies with alternative sample and loan

specifications used in the analysis. Note that I focus on loan facility per year.

4. RESULTS

4.1. Descriptive Statistics

Table 1 provides descriptive statistics for the total sample. The mean and median levels for organizational

capital stock makes up 70% to 101% of the total book value of assets for an average firm for the total Compustat

sample. These levels are very close to those observed in the past research (e.g., Francis, Mani and Wu, 2015).

An average firm pays 125BPS (median)/155BPS (mean) in additional spreads over LIBOR with 5 (median)/4

(mean) year duration for a typical loan. On average, a typical loan includes 8 (mean) unique restrictive loan

covenants. Table 2 test for mean level differences in some of the variables modelled in the analysis. Panel A,

Panel B and Panel C presents results for high and low levels (all based on median) of institutional ownership,

analyst coverage and state level non-compete enforceability. Under the univariate setting, investment in

organizational capital seems to be levelled for firms with high/low institutional ownership and analyst coverage.

This observation, however, turns positive under multivariate analysis (un-tabulated) for firms with better

corporate governance. On the other hand, loan spreads, loan size, number of restrictive covenants all have

predicted signs in favor of firms with stronger corporate governance. Likewise, firms with stronger corporate

governance are larger (LNTA) and more profitable (ROA) in comparison to firms with weaker corporate

governance. Given these cross-sectional differences in alternative sample specifications used in the analysis, we

apply propensity score matching to better identify sample-groups that more closely represent homogenous sub-

samples.

Interestingly Panel C shows that firms that operate in states with stricter applications of non-compete laws

(NCEI>M) invest significantly more in organizational capital compared to those who do not (NCEI<M). This

observation is also confirmed in un-tabulated multivariate analysis which is in line with the general findings that

firms in these states do tend to invest in their human capital (Starr, 2016). Although not observed in multivariate

analysis (un-tabulated) firms that operate in states with stricter NCAs have lower loan spreads compared those

who do not. Moreover, the analysis indicates that investment in tangible and fixed capital as well as R&D is

higher/(lower) for firms that operate in states with loose/(strict) application of NCA rules. This observation

remains unchanged when the analysis use Garmaise (2009) NCA index which limits the sample into years prior

to 1994. Nonetheless, one should note that our analysis is more focused at understanding creditors’ perception

of risks associated with intangible-based capital investments. Table 3 presents Pearson (below diagonal) and

Spearman (above diagonal) correlations among observable firm characteristics. The table indicates no particular

link between OCAP and other firm characteristics.

4.2. Organizational Capital and Loan Spreads

Table 4 examines the link between one-period lagged organizational capital and bank loan spreads e.q. (3) using

both total and propensity-score-matched sub-samples. Using the total sample, columns 2 and 3 show that both

firm level (OC) and industry-adjusted organizational capital is associated with lower loan spreads. Next, in

columns 4 and 5 I include fixed capital (CAPX) and R&D investments into the model. This analysis allows me

to examine, from the perspective of lending institutions, the riskiness (if any) of different investment alternatives

and how intangible-based OC relates to it. While I continue to observe negative links between both types of

OCAP and loan spreads I fail to find any directional link between other types of investments and cost of bank

financing. In economic terms, a standard deviation increase in organizational capital stock helps reduce loan

spreads by 4.70BPS on average which results in $1.14MN savings over the average loan period (4 years). In un-

tabulated results I also estimate accumulated stocks for R&D investments using the same methodology and

assumptions made estimating the OCAP. My results are robust to the inclusion of proxy variable that capitalizes

R&D investments using the same procedure to capitalize investment in OCAP.

I observe predicted signs for other control variables. For example, on average loans with performance pricing

provisions (PPP), larger lead (NLD) and non-lead (NPARTD) syndicate formations, larger lead-level ownership

(LPCT) and those including the most reputable lenders (LREP) are associated with lower loan spreads on

average. Similarly, I observe a negative link between revolving credit facilities, which proxies for relationship-

focused lending (e.g., Berger and Udell, 1995), and loan spreads. On the other hand, covenant intensity (NCOV)

and loan term (TERM) increase loan spreads. Moreover, I observe negative and statistically significant

coefficients for firm (LNTA) and loan size (LNLOAN), indicating the positive role of scale economies play in

bank financing. The analysis also yields negative coefficients for greater profitability (PTROA, PIFO), earnings

quality, (AQ) and stability (SDCFO). In line with the past research, both the level of financial leverage (LVRG)

and the proxy for financial health (ZSCORE) have the predicted signs. Finally, I find positive links between

both types proxies of macro-level financial constraints (CIS, CFNAI) and loan spreads.10

Columns 6 to 9 present the results for the analysis that score matches firm-level observable characteristics to

leave out investment in OCAP stock as a treatment effect. This analysis controls for confounding effects other

variables might have on the observed link between OCAP and loan spreads. The details of the probit regression

used and the balancing properties are provided in Appendix C. Columns 6 and 7 confirm earlier observations

that accumulated stock of OCAP is negatively associated with loan spreads. Columns 8 and 9 replace the

continuous measure of OCAP stock with a dummy indicator (HOCAP) that is 1/(0) for OC and IAOC variables

greater/(less) than sample median. Results remain robust to the replacement of the continuous OCAP measure

with a binary variable. Overall, these results are in line with the predictions made in H1.

4.3. Organizational Capital and Loan Spreads – Controlling for External Governance

Table 5 controls for the effects of corporate governance measures on the contractual terms associated with

organizational capital. Panel A presents results for institutional ownership levels and Panel B presents results for

the level of analyst coverage. Each panel segregates the sample based on above/below median level of measures

used to proxy for corporate governance. Focusing on the total sample in both panels, the results show that

organizational capital is negatively associated with loan spreads and that this link is economically and

statistically stronger for firms with higher institutional ownership (HOWN) and analyst coverage (HAFOL).

Specifically, a standard deviation increase in organizational capital reduces loan spreads by 10BPS/(11BPS) for

firms with above median institutional ownership/(analyst coverage), which translates into around

$2.50/($2.80MN) savings over the 4-year average loan term. These results are robust to a model specification

(un-tabulated) where we incorporate governance proxies within a model and use interaction analysis to test the

effects of corporate governance on our results. These results and interpretations are qualitatively confirmed

using propensity-score-matched sub-samples for both types of governance proxies.

4.4. Organizational Capital and Loan Spreads – Controlling for State Level NCAs

Table 6 uses total (Panel A) and score-matched samples (Panel B) to test for whether state-level NCAs affect

creditors’ perception of business risk related to intangible based OCAP. In each panel, I divide the sample into

sub-parts based on state-level NCA enforceability measure. Accordingly NCEI>MED/( NCEI<MED) represents

10 Note that for brevity’s sake I report only the former in the tables.

sub-samples in states with above/(below) median level NCEI, hence strong/(weak) state-level NCA

applicability. Panel B also use the same sample-segregation strategy except that the sample itself is propensity-

score-matched to control for firm level characteristics that might confound the intersection between OCAP,

state-level NCA intensity and bank cost of financing. In both panels the results show that creditors charge lower

spreads for investment in both firm-specific (OC) and industry-adjusted (IAOC) levels of OCAP stock for firms

that operate in states with stricter NCA applications (NCE>MED or NCE>M) in comparison to those made by

firms that operate in states with looser NCA applications. In economic terms a standard deviation increase in

investment in OCAP stock reduces loan spreads by 8.30BPS for firms that operate in states with stricter NCA

rules as opposed to 6.40BPS reduction for firms that operate in states with looser NCA rules. Importantly,

however, this spread difference is not statistically robust enough to support anticipated benefits from state-level

frictions that aims to protect knowledge-based capital from migrating to competing firms via employment

exchange. I find qualitatively similar results when state-level non-compete intensity (NCEI) is modelled within

the analysis and interacted with organizational capital. These results reject the hypothesis 2 and are robust to

using alternative score-matching techniques and caliper adjustments.

4.5. Organizational Capital and Non-Price Loan Terms

Table 7 examines the link between investments in OCAP and non-price loan terms using both total and matched

samples. Banks may choose to account for risks associated with heavy reliance on intangible-based capital using

non-price terms rather than price terms. This is an intuitive presumption given that lenders, depending on the

market circumstances, may trade certain non-price loan terms off for additional yield or vice versa (Stein, 2013).

Accordingly, lenders can use restrictive covenant clauses and/or performance pricing provisions with the aim to

mitigate borrower-lender frictions (e.g., Nini et al., 2009; Demiroglu and James, 2010) and apply stricter forms

of non-price terms to the extent that they perceive heavy reliance on intangible-based investments as more risky

relative to investments in tangible assets. Results using both total and matched sub-samples indicate that

investment in OCAP is associated with lower number of restrictive covenants (LNCOV) and facilitate access to

larger lending facilities (LNLOAN). Notably, however, under both sample specifications R&D intensity is

associated with significantly smaller lending facilities. These results indicate that lenders are likely to perceive

R&D intensive firms relatively riskier, given the uncertainties associated with expected cash flows from such

investments. On the other hand, I fail to observe any relation between asset tangibility and covenant intensity

and/or loan size.

Next, in Table 8 I examine the link between non-price loan terms and OCAP stock controlling for corporate

governance. Neither panels provide a statistically definitive cross-sectionality in non-price loan terms associated

with organizational capital. Results generally confirm the baseline evidence (Table 7) that, on average,

investment in organizational capital enables firms to have access to larger loan facilities with less strict covenant

structures.

Finally, Table 9 examines the link between non-price loan terms and OCAP stock controlling for state-level

NCAs using a matched sample. Using the total sample in Panel A and the matched sample in Panel B results

indicate a negative link between organizational capital and the number of restrictive loan covenants for firms

that operate in states with stricter NCA enforceability. On the other hand, I observe no such relation for firms

that operate in states with employee-friendly non-compete laws. This difference among the two alternative state-

groups are statistically significant at p<0.0001. These results indicate that lender seem to reflect some of the

cash flow ownership risks associated with knowledge-based capital via stricter covenant terms. Moreover, with

the exception of the results in Panel B, investment in OCAP stock is positively related to the size of the

syndicate loan facilities for firms that operate both in states with stricter and looser applications of NCA rules.11

As in Table 7 and Table 8, I fail to observe a robust and a directionally-definitive link between asset tangibility

and loan size. I continue to observe a negative link between R&D intensity and loan size albeit the evidence

seems more pronounced for firms that operate in states with looser NCA rules. Accordingly, the analysis

indicate that potential competitive threats from knowledge-leakage might extent towards R&D based

investments when firms operate in states that attain more liberal employee mobility policies.12

5. CONCLUSION

This paper investigates the direct link between investment in organizational capital, the most important

multiplier of firm-level intangible capital, and price and non-price terms associated with loan financing. In line

with the role intellectual capital plays in value creation in “the new economy” (e.g., Atkeson and Kehoe, 2005;

11 Note that in un-tabulated analysis when the test variable OCAP is exchanged with HOCAP, a dummy

indicator that takes the value of 1 if firm-level OCAP investment is larger than the sample median, the

coefficients in Panel B becomes statistically significant for firms that operate in states with stricter NCA rules

(NCEI>MED). I observe similarly positive and statistically significant link between loan size and firm-level

investment in OCAP stocks at alternative sample distribution points including firms with OCAP investment at

the top quintile, quartile and decile. 12 I do not, in any part of this paper, aim to measure R&D investment spillovers, neither from firms’ nor from

lenders’ perspective. Accordingly, I do apply a degree of caution in interpreting these results. Nonetheless, more

lender-focused research seems necessary on the topic in order to provide more robust and definitive explanation.

Hulten and Hao, 2008; Corrado, Hulten and Sichel, 2009; Corrado and Hulten, 2010; Fu, Huang and Wang,

2015; OECD, 2013; Lev et al., 2016) lenders seem to acknowledge corporate-wide switch towards hard-to-value

investments in intellectual capital over time and the associated benefits thereof. Using alternative sample and

model specifications, including propensity score matching techniques to establish more robust casual inference,

I find that investment in organizational capital is associated with favourable price and non-price loan terms on

average. In additional analysis, I find that these favourable price-based loan terms associated with organizational

capital are more pronounced for firms with stronger corporate governance. And finally I find neither economical

nor statistical difference in price based terms associated with organizational capital for firms that operate in

states that apply stricter non-compete employment agreements, sate-level laws that are argued to help shield

migration of knowledge-based capital (if any) via key-talent outside options, compared to those that do not.

These results have important implications for academic research and policy makers on several accounts. First,

unlike Loumioti (2012) and Lim et al. (2016) that use intangibles that are capitalized, predominantly as part of

business combinations and/or acquisitions, I focus on intangible capital that is internally generated and expensed

due to accounting restrictions. In particular, results show that the economic competencies achieved via

knowledge-based capital investments extend beyond the collateral value of intangible assets from lenders’

perspective – complementing the evidence Loumioti (2012). Moreover, somewhat deviating from the arguments

made in Loumioti (2012) and Lim et al. (2016), the ability to access to larger financing facilities with looser

covenant structures indicates that lenders do not perceive firms that heavily rely on internally-generated

knowledge-based capital as riskier, on average, in comparison to firms with larger investments in tangible

capital. These results extend beyond arguments that historically focused on asset tangibility and firm-level

flexibility in accessing alternative sources of capital (e.g., Almeida and Campello, 2007; Campello and

Giambiona, 2010). Moreover, the inferences are timely and relevant in the light of systematical increase in

intangible capital and ongoing debate as to whether and to what extend these investments should be capitalized

(e.g., Lev and Sugiannis, 1996; Kothari, Laguerre and Leone, 2002; Kanodia et al., 2004; Skinner, 2008; Lev,

2008; Zéghal and Maaloul, 2011). Notably, I document that while lenders rationally assess value-generation

from organizational capital as a form of intangible capital as indicated in Skinner (2008), I fail to document such

contractual efficiencies for internally-generated intangible capital stemming from R&D investments/capital

stock.

Finally, unlike the case for shareholders (Eisfeldt and Papanikolaou, 2013; 2014), lenders do not seem to

acknowledge potential cash flow ownership risks associated with intellectual capital stock, at least not to an

extent that one would expect given the prevalence of these state-level employment frictions. A number of

alternative factors can influence our observations. First, lenders might pay subordinate attention to these

provisions which requires deeper investigation to have more substantial argument on the case. Second, lenders

may perceive the benefits from investing in organizational capital to be embedded in firm-specific knowledge,

systems and procedures that are hard-to-imitate by competitors and remain so regardless of the key-talent

outside options (e.g., Atkeson and Kehoe, 2005; Lev et al., 2009; Lev et al. 2016). Third, in essence non-

compete agreements only delay skill-transfer and are subject to geographic and/or time-specific limitations

which may render these employment restrictions as ineffective (eventually) in the long run, at least from the

lenders’ point of view. Finally, one may also argue that the sample coverage is inevitably biased towards larger

firms and may not reflect genuine benefits associated with non-compete agreement that might accrue to smaller

firms/units, a topic that also warrants further attention. Nonetheless, I believe the inferences from the analysis

renders some credit to arguments made in regulatory statements and popular press challenging the effectiveness

of non-compete agreements (e.g., DoT 2016, Whitehouse, 2016; e.g., Jasper, 2010; Muro, 2016; Viswanatha,

2016; Ben-Shahar, 2016) and relate to the recent evidence that links competitive threats on knowledge-based

capital and conservative capital structures (Klasa et al., 2016).

6. REFERENCES

Almeida, H., & Campello, M. (2007). Financial Constraints, Asset Tangibility, and Corporate Investment. Review of

Financial Studies, 20(5), 1429–1460. https://doi.org/10.1093/rfs/hhm019

Asquith, P., Beatty, A., & Weber, J. (2005). Performance pricing in bank debt contracts. Journal of Accounting and

Economics 40, 101–128. doi:10.1016/j.jacceco.2004.09.005

Atkeson, A., & Kehoe, P.J. (2005). Modeling and Measuring Organization Capital. Journal of Political Economy,

113(5), 1026–1053. https://doi.org/10.1086/431289

Ball, R., Bushman, R.M., Vasvari, F.P. (2008). The Debt-Contracting Value of Accounting Information and Loan

Syndicate Structure. Journal of Accounting Research 46, 247–287. doi:10.1111/j.1475-679X.2008.00273.x

Benmelech, E., (2009). Asset Salability and Debt Maturity: Evidence from 19th Century

American Railroads. Review of Financial Studies 22, 1545-1583.

Benmelech, E., and Bergman, N. (2009). Collateral Pricing. Journal of Financial Economics

91, 339-360.

Ben-Shahar. (2016). Obama’s Pitch to Ban Non-Compete Agreements Would Make the Rich Richer. Retrieved from

http://fortune.com/2016/11/03/obama-non-compete-agreements/

Berger, A.N., & Udell, G.F. (1995). Relationship Lending and Lines of Credit in Small Firm Finance. The Journal of

Business 68, 351–381.

Blair, M. M., & Wallman, S. M. H. (2000). Unseen Wealth: Report of the Brookings Task Force on Intangibles.

Brookings Institution Press.

Boguth, O., Newton, D., & Simutin, M. (2016). The Fragility of Organization Capital. Available at SSRN 2784425.

Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2784425

Campello, M., & Giambona, E., (2010). Capital structure and the redeployability of tangible assets. Retrieved from

http://dare.uva.nl/record/1/328884

Chava, S., & Roberts, M.R. (2008). How Does Financing Impact Investment? The Role of Debt Covenants. The

Journal of Finance 63, 2085–2121.

Chen, W., & Inklaar, R. (2015). Productivity spillovers of organization capital. Journal of Productivity Analysis, 45(3),

229–245. https://doi.org/10.1007/s11123-015-0463-x

Corrado, C., Haltiwanger, J.C., &Sichel, D.E. (2005). Conference on Research in Income and Wealth (Eds.).

Measuring capital in the new economy. Chicago: University of Chicago Press.

Corrado, C., Hulten, C., & Sichel, D. (2009). Intangible Capital and U.S. Economic Growth. Review of Income and

Wealth, 55(3), 661–685. https://doi.org/10.1111/j.1475-4991.2009.00343.x

Corrado, C.A., & Hulten, C.R. (2010). How Do You Measure a “Technological Revolution”? The American Economic

Review, 100(2), 99–104.

Corrado, C., Haskel, J., Jona-Lasinio, C., & Iommi, M. (2013). Innovation and intangible investment in Europe, Japan,

and the United States. Oxford Review of Economic Policy, 29(2), 261–286. https://doi.org/10.1093/oxrep/grt017

Corrado, C., Haskel, J., Jona-Lasinio, C., & Iommi, M. (2014). Intangibles and Industry Productivity Growth: Evidence

from the EU. IARIW 33rd General Conference Rotterdam, the Netherlands, August 24-30, 2014.

Dennis, S.A., Mullineaux, D.J., (2000). Syndicated Loans. Journal of Financial Intermediation 9, 404–426.

doi:10.1006/jfin.2000.0298

Eckstein, C. (2004). The measurement and recognition of intangible assets: then and now. Accounting Forum, 28(2),

139–158. https://doi.org/10.1016/j.accfor.2004.02.001

Eisfeldt, A. L., & Papanikolaou, D. (2013). Organization Capital and the Cross-Section of Expected Returns. The

Journal of Finance, 68(4), 1365–1406. https://doi.org/10.1111/jofi.12034

Eisfeldt, A. L., & Papanikolaou, D. (2014). The Value and Ownership of Intangible Capital. The American Economic

Review, 104(5), 189–194. https://doi.org/10.1257/aer.104.5.189

E&Y (2011). US GAAP vs. IFRS. Retrieved from

http://www.ey.com/Publication/vwLUAssets/US_GAAP_v_IFRS:_The_Basics/$FILE/US%20GAAP%20v%20

IFRS%20Dec%202011.pdf

Francis, B. B., Mani, S. B., & Wu, Q. (2015). The Impact of Organization Capital on Firm Innovation. Available at

SSRN. Retrieved from http://papers.ssrn.com/sol3/Papers.cfm?abstract_id=2675779

Fu, F., Huang, S., & Wang, R. (2015). Why Do US Firms Invest Less Over Time? Available at SSRN 2564451.

Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2564451

Garmaise, M. J. (2009). Ties that truly bind: Noncompetition agreements, executive compensation, and firm

investment. Journal of Law, Economics, and Organization. Retrieved from

http://jleo.oxfordjournals.org/content/early/2009/11/03/jleo.ewp033.short

Gilson, R. J. (1999). The Legal Infrastructure of High Technology Industrial Districts: Silicon Valley, Route 128, and

Covenants Not to Compete. New York University Law Review, 74, 575.

Gillan, S., J. Hartzell & Parrino, R. (2005). Explicit vs. Implicit Contracts: Evidence from CEO

Employment Agreements, Working Paper.

Harford, J., Klasa, S., Maxwell, W.F. (2014). Refinancing Risk and Cash Holdings. The Journal of Finance 69, 975–

1012. doi:10.1111/jofi.12133

Hart, O., & Moore, J. (1994). A theory of Debt Based on the Inalienability of Human Capital. Quarterly Journal of

Economics 109, 841-879.

Holmstrom, B., & Jean, T. (1997). Financial Intermediation, Loanable Funds, and the Real Sector. Quarterly Journal of

Economics 3, 663-691.

Hulten, C. R., & Hao, X. (2008). What is a company really worth. Intangible Capital and the’Market to Book

Value’Puzzle, NBER Wp, 14548. Retrieved from

http://raw.rutgers.edu/docs/intangibles/Presentations/Rutgers%20Conference%20Sept%202010%20Janet%20H

ao%202.pdf

Hulten, C., Hao, J., & Jaeger, K. (2010). Macro versus Micro Comparisons of Intangible Capital: The Case of

Germany and the US. CoInvest publications. Retrieved from

http://eeclab.org.uk/pub/CoInvest/COINVESTHulteninterim/COINVEST_HHJ_FINAL-1.pdf

Jasper, E. (n.d.). Putting You in Handcuffs: The Non-Compete Agreement. Retrieved November 27, 2016, from

http://www.forbes.com/sites/work-in-progress/2010/09/14/putting-you-in-handcuffs-the-non-compete-

agreement/

Kaplan, S. N., & Strömberg, P. (2003). Financial Contracting Theory Meets the Real World: An Empirical Analysis of

Venture Capital Contracts. The Review of Economic Studies, 70(2), 281–315. https://doi.org/10.1111/1467-

937X.00245

Kanodia, C., Sapra, H., & Venugopalan, R. (2004). Should Intangibles Be Measured: What Are the Economic Trade-

Offs? Journal of Accounting Research, 42(1), 89–120. https://doi.org/10.1111/j.1475-679X.2004.00130.x

Lev, B., & Zarowin, P. (1999). The Boundaries of Financial Reporting and How to Extend Them (Digest Summary).

Journal of Accounting Research, 37(2), 353–385.

Lev, B., & Daum, J. H. (2004). The dominance of intangible assets: consequences for enterprise management and

corporate reporting. Measuring Business Excellence, 8(1), 6–17. https://doi.org/10.1108/13683040410524694

Lev, B., & Radhakrishnan, S. (2005). The valuation of organization capital. In Measuring capital in the new economy:

University of Chicago Press, 73-110.

Lev, B. (2008). A rejoinder to Douglas Skinner’s “Accounting for intangibles – a critical review of policy

recommendations.” Accounting and Business Research, 38(3), 209–213.

https://doi.org/10.1080/00014788.2008.9663334

Lev, B., Radhakrishnan, S., & Zhang, W. (2009). Organization Capital. Abacus, 45 (3), 275-298.

Lim, S. C., Macias, A. J., & Moeller, T. (2015). Intangible Assets and Capital Structure. Available at SSRN 2514551.

Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2514551

Long, B. S. (2005). Protecting Employer Investment in Training: Noncompetes vs. Repayment Agreements. Duke Law

Journal, 54(5), 1295–1320.

Loumioti, M. (2012). The use of intangible assets as loan collateral. Available at SSRN 1748675. Retrieved from

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1748675

Marx, M., Strumsky, D., & Fleming, L. (2009). Mobility, Skills, and the Michigan Non-Compete Experiment.

Management Science, 55(6), 875–889. https://doi.org/10.1287/mnsc.1080.0985

Monzur, M. H. (2015). Three essays on corporate finance. Retrieved from http://espace.library.curtin.edu.au/cgi-

bin/espace.pdf?file=/2015/10/12/file_1/229659

Mora, N., (2015). Lender Exposure and Effort in the Syndicated Loan Market. Journal of Risk and Insurance 82, 205–

252. doi:10.1111/jori.1202

Muro, M. (2016, May 23). Why Noncompete Pacts Are Bad for Workers–and the Economy. Retrieved from

http://blogs.wsj.com/experts/2016/05/23/why-states-should-stop-the-spread-of-noncompete-pacts/

Nini, G., Smith, D. C., & Sufi, A. (2009). Creditor control rights and firm investment policy. Journal of Financial

Economics, 92(3), 400–420. ttps://doi.org/10.1016/j.jfineco.2008.04.008

OECD. (2013). New Sources of Growth: Knowledge-based Capital Driving Investment and Productivity in the 21st

century -Interim Project Findings. OECD, Paris

Rauh, J. D., & Sufi, A. (2010). Capital Structure and Debt Structure. Rev. Financial Studies. hhq095.

doi:10.1093/rfs/hhq095

Rajan, R., & Winton, A., (1995). Covenants and Collateral as Incentives to Monitor. The Journal of Finance 50, 1113–

1146. doi:10.2307/2329346

Roberts, G., & Yuan, L. (Edward). (2010). Does institutional ownership affect the cost of bank borrowing? Journal of

Economics and Business, 62(6), 604–626. https://doi.org/10.1016/j.jeconbus.2009.05.002

Samila, S., & Sorenson, O. (2011). Noncompete Covenants: Incentives to Innovate or Impediments to Growth.

Management Science, 57(3), 425–438. https://doi.org/10.1287/mnsc.1100.1280

Shleifer, A., & Vishny, R. W. (1992). Liquidation Values and Debt Capacity: A Market Equilibrium Approach. The

Journal of Finance, 47(4), 1343–1366. https://doi.org/10.1111/j.1540-6261.1992.tb04661.x

Skinner, D. J. (2008). Accounting for intangibles – a critical review of policy recommendations. Accounting and

Business Research, 38(3), 191–204. https://doi.org/10.1080/00014788.2008.9663332

Stein, J. (2013). Remarks at the Restoring Household Financial Stability after the Great Recession Research

Symposium, Federal Reserve Bank of St. Louis. Retrieved from

https://www.federalreserve.gov/newsevents/speech/stein20130207a.htm

Sufi, A., 2007. Information Asymmetry and Financing Arrangements: Evidence from Syndicated Loans. The Journal of

Finance 62, 629–668. doi:10.1111/j.1540-6261.2007.01219.x

Starr, E., Prescott, J.J., & Bishara, N. (2016a). Noncompetes in the U.S. Labor Force. Working Paper.

Starr, E. (2016). Consider This: Training, Wages, and the Enforceability of Covenants Not to Compete. Working

Paper.

Starr, E., Ganco, M., & Benjamin, C. (2016). Redirect and Retain: How Firms Capitalize on Within and Across

Industry Mobility Frictions. Working Paper.

DoT. (2016). Non-compete Contracts: Economic Effects and Policy Implications

Viswanatha, A. (2016, February 2). Noncompete Agreements Hobble Junior Employees. Wall Street Journal.

Retrieved from http://www.wsj.com/articles/noncompete-agreements-hobble-junior-employees-1454441651

West, J. K. (2015). Synthesis of Enquiries into Intellectual Property’s Economic Impact’. Retrieved from

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2652044

White House. (2016). Non-Compete Agreements: Analysis of the Usage, Potential Issues, and State Responses

Younge, K. A., & Marx, M. (2016). The Value of Employee Retention: Evidence From a Natural Experiment. Journal

of Economics & Management Strategy, 25(3), 652–677. https://doi.org/10.1111/jems.12154

Zéghal, D., & Maaloul, A. (2011). The accounting treatment of intangibles – A critical review of the literature.

Accounting Forum, 35(4), 262–274. https://doi.org/10.1016/j.accfor.2011.04.003

Table 1

Summary Statistics

Variables Q1 Med Mean Q4 Sdev

Organizational Capital

𝑂𝐶 0.33 0.70 1.01 1.29 1.20

𝐼𝐴𝑂𝐶 -0.54 -0.19 -0.00 0.20 1.11

𝑂𝐶𝑄 2 3 3 4 1.41

Loan Variables

𝑆𝑃𝑅𝐸𝐴𝐷 (in basis) 62.5 125 155 225 125

𝑁𝑂𝐿𝐸𝐴𝐷 1 3 3 4 2.09

𝑁𝑂𝑃𝐴𝑅𝑇 3 6 8 10 7.66

𝐿𝐸𝐴𝐷𝑃𝐶𝑇 0.20 0.33 0.45 0.75 0.34

𝐿𝑅𝐸𝑃 0 1 0.66 1 0.47

𝑁𝑈𝑀𝐶𝑂𝑉 0 0 8.39 5 20.44

𝑇𝐸𝑅𝑀 3 5 3.98 5 1.69

𝐿𝑁𝐿𝑂𝐴𝑁 4.61 5.52 5.58 6.46 1.31

𝑅𝐸𝑉𝐷 1 1 .78 1 0.42

𝑃𝑃𝑃 0 0 0.33 1 0.47

𝑆𝐸𝐶𝑈𝑅 0 0 0.39 1 0.49

Other Variables

𝑃𝐼𝐹𝑂 0 0 0.01 0.02 0.04

𝐶𝐼𝑆 1.69 1.92 1.99 2.35 0.51

𝐶𝐹𝑁𝐴𝐼 -0.43 0.07 -0.11 0.35 1.01

𝑍𝑆𝐶𝑂𝑅𝐸 1.80 2.69 3.31 4.02 3.31

𝑆𝐷𝐶𝐹𝑂 0.04 0.05 0.06 0.07 0.05

𝑃𝑇𝑅𝑂𝐴 0.013 0.06 0.05 0.11 0.18

𝐿𝑁𝑇𝐴 6.10 7.13 7.23 8.27 1.61

𝑇𝐴𝑁𝐺 0.076 0.18 0.26 0.39 0.24

𝑅𝑁𝐷 0 0 0.06 0.063 0.144

𝐶𝐴𝑃𝑋 0.017 0.036 0.06 0.072 0.078

𝐿𝑉𝑅𝐺 0.13 0.25 0.28 0.39 0.22

𝐴𝑄 0.03 0.05 0.07 0.08 0.11

𝐴𝐹𝑂𝐿 2 6 8.19 12 7.68

𝐼𝑂𝑊𝑁 0.00 0.45 0.51 0.77 0.38

𝑀𝑇𝐵 1.24 2.06 4.62 3.42 104.71

Notes:

Table 1 presents summary statistics. Q1 and Q4 represent the bottom and top quartiles for each

observation. All variables are explained in greater detail in Appendix A.

Table 2

Cross-Sectional Differences in Variables

Panel A: Institutional Ownership

Variables HOWN LOWN diff p-value

𝑂𝐶 1,07 1.08 -0.01 0.81

𝑆𝑃𝑅𝐸𝐴𝐷 (in basis) 123.56 187 -63.44 0.00***

𝑁𝐶𝑂𝑉 5.88 10.88 -5.00 0.00***

𝐿𝑁𝐿𝑂𝐴𝑁 785 458 327 0.00***

𝑇𝐴𝑁𝐺 0.29 0.35 -0.06 0.00***

𝑅𝑁𝐷 0.018 0.016 0.002 0.02**

𝐶𝐴𝑃𝑋 0.053 0.08 -0.027 0.00***

𝑃𝑇𝑅𝑂𝐴 0.082 0.01 0.072 0.00***

𝐿𝑉𝑅𝐺 0.23 0.32 -0.09 0.00***

𝐿𝑁𝑇𝐴 7.82 6.63 1.19 0.00***

𝑍𝑆𝐶𝑂𝑅𝐸 3.59 3.01 0.58 0.00***

Panel B: Analysts Following

Variables HAFOL LAFOL diff p-value

𝑂𝐶 0.80 1.26 0.14 0.00***

𝑆𝑃𝑅𝐸𝐴𝐷 (in basis) 126 188 -62 0.00***

𝑁𝐶𝑂𝑉 5.72 11.43 -5.71 0.00***

𝐿𝑁𝐿𝑂𝐴𝑁 6.00 5.00 1.00 0.00***

𝑇𝐴𝑁𝐺 0.33 0.31 0.02 0.00***

𝑅𝑁𝐷 0.019 0.014 0.005 0.00***

𝐶𝐴𝑃𝑋 0.069 0.063 -0.006 0.00***

𝑃𝑇𝑅𝑂𝐴 0.069 0.019 0.05 0.00***

𝐿𝑉𝑅𝐺 0.26 0.30 -0.04 0.00***

𝐿𝑁𝑇𝐴 6.34 8.00 -0.66 0.00***

𝑍𝑆𝐶𝑂𝑅𝐸 3.51 3.07 0.44 0.00***

Panel C: Concentrated Customers

Variables NCEI>M NCEI<M diff p-value

𝑂𝐶 1.06 0.97 0.085 0.00***

𝑆𝑃𝑅𝐸𝐴𝐷 (in basis) 150 160 -10 0.00

𝑁𝐶𝑂𝑉 7.69 7.30 0.38 0.24

𝐿𝑁𝐿𝑂𝐴𝑁 5.44 5.40 0.05 0.05*

𝑇𝐴𝑁𝐺 0.25 0.28 -0.03 0.00***

𝑅𝑁𝐷 0.05 0.057 -0.014 0.00***

𝐶𝐴𝑃𝑋 0.053 0.065 -0.012 0.00***

𝑃𝑇𝑅𝑂𝐴 -0.03 0.057 0.027 0.00***

𝐿𝑉𝑅𝐺 0.21 0.20 0.01 0.00***

𝐿𝑁𝑇𝐴 5.56 5.40 0.16 0.00***

𝑍𝑆𝐶𝑂𝑅𝐸 6.05 7.11 -1.06 0.03***

Notes: Table 2 presents mean-level differences in observable characteristics for firms with

stronger/(weaker) corporate governance (Panel A and Panel B) as well as for firms that operate in

states with stricter (NCEI>M) and looser (NCEI>M) NCA intensity (Panel C). Asterisks above the

coefficients represent significance levels where * is used for p < 10%, ** is used for p < 5% *** is

used for p < 1% significance levels. All variables are explained in greater detail in Appendix A.

Table 3

Correlation Structure of the Variables

(1) (2) (3) (4) (5 (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17)

OC (1) 1.000 -0.123 0.064 -0.103 -0.198 -0.108 0.0623 0.121 0.019 -0.141 -0.029 0.402 0.114 -0.216 -0.272 -0.197 0.102

. 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.061 0.000 0.004 0.000 0.000 0.000 0.000 0.000 0.000

SPREAD (2) -0.017 1.000 -0.134 0.091 -0.231 -0.335 -0.216 0.086 -0.272 0.390 -0.074 -0.321 -0.40 -0.297 0.213 -0.019 0.209 0.100 . 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.072 0.000

NCOV (3) 0.007 -0.008 1.000 -0.066 -0.172 -0.193 -0.188 -0.024 -0.151 -0.397 -0.035 -0.037 -0.067 -0.260 0.105 -0.031 0.039 0.472 0.426 . 0.000 0.000 0.000 0.000 0.019 0.000 0.000 0.000 0.003 0.000 0.000 0.000 0.002 0.000

TERM (4) -0.071 0.088 0.033 1.000 0.227 -0.043 0.043 -0.064 0.013 0.087 0.132 -0.043 0.025 0.046 0.169 -0.045 -0.105

0.000 0.000 0.001 . 0.000 0.000 0.000 0.000 0.192 0.000 0.000 0.000 0.015 0.000 0.000 0.000 0.000

𝐿𝑁𝐿𝑂𝐴𝑁 (5) -0.126 -0.189 -0.109 0.177 1.000 0.463 0.248 -0.117 0.336 0.192 0.042 -0.119 0.101 0.795 0.169 0.032 -0.274

0.000 0.000 0.000 0.000 . 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.000

𝐴𝐹𝑂𝐿 (6) -0.083 -0.279 -0.152 -0.076 0.471 1.000 0.269 -0.077 0.293 0.103 0.086 0.147 0.265 0.601 -0.088 0.035 -0.122

0.000 0.000 0.000 0.000 0.000 . 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

𝐼𝑂𝑊𝑁 (7) -0.234 -0.141 0.009 0.257 0.237 0.005 1.000 0.312 0.218 0.027 0.205 0.277 0.338 -0.178 -0.051 -0.193

0.157 0.000 0.000 0.350 0.000 0.000 . 0.638 0.000 0.000 0.007 0.000 0.000 0.000 0.000 0.000 0.000

𝐴𝑄 (8) -0.006 0.036 -0.024 -0.015 -0.047 -0.034 -0.004 1.000 0.023 0.055 -0.029 0.097 -0.042 -0.113 -0.150 -0.212 0.341

0.537 0.000 0.018 0.137 0.000 0.000 0.665 . 0.021 0.000 0.004 0.000 0.000 0.000 0.000 0.000 0.000

𝑃𝐼𝐹𝑂 (9) -0.023 -0.089 -0.043 -0.067 0.172 0.153 0.049 0.000 1.000 0.128 0.028 0.144 0.281 0.445 -0.118 -0.107 -0.193

0.025 0.000 0.000 0.000 0.000 0.000 0.000 0.950 . 0.000 0.007 0.000 0.000 0.000 0.000 0.000 0.000

𝐶𝐼𝑆 (10) -0.095 0.346 -0.254 0.096 0.185 0.109 0.210 0.049 0.038 1.000 -0.059 -0.046 -0.018 0.253 -0.034 0.005 -0.008

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 . 0.000 0.000 0.072 0.000 0.000 0.594 0.406

𝐶𝐹𝑁𝐴𝐼 (11) -0.003 -0.086 0.036 0.094 0.045 0.045 -0.012 -0.012 -0.013 -0.073 1.000 0.071 0.103 0.015 -0.009 -0.017 -0.024

0.723 0.000 0.000 0.000 0.000 0.000 0.223 0.223 0.195 0.000 . 0.000 0.000 0.134 0.345 0.093 0.019

𝑍𝑆𝐶𝑂𝑅𝐸 (12) 0.269 -0.183 -0.052 -0.038 -0.115 0.100 0.088 0.030 0.018 -0.049 0.062 1.000 0.665 -0.104 -0.537 -0.202 0.115

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.073 0.000 0.000 . 0.000 0.000 0.000 0.000 0.000

𝑃𝑇𝑅𝑂𝐴 (13) -0.020 -0.263 -0.045 0.035 0.085 0.163 0.215 -0.014 0.077 -0.001 0.109 0.375 1.000 0.120 -0.285 -0.062 -0.063

0.043 0.000 0.000 0.000 0.000 0.000 0.000 0.168 0.000 0.901 0.000 0.000 . 0.000 0.000 0.000 0.000

𝐿𝑁𝑇𝐴 (14) -0.172 -0.244 -0.204 -0.020 0.797 0.623 0.341 -0.041 0.2518 0.235 0.023 -0.107 0.122 1.000 0.071 0.049 -0.316

0.000 0.000 0.000 0.052 0.000 0.000 0.000 0.000 0.000 0.000 0.026 0.000 0.000 . 0.000 0.000 0.000

𝐿𝑉𝑅𝐺 (15) -0.125 0.226 0.164 0.186 0.139 -0.124 -0.212 -0.058 -0.059 -0.037 0.017 -0.3541 -0.158 0.017 1.000 0.147 -0.150

0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.096 0.000 0.000 0.085 . 0.000 0.000

𝑇𝐴𝑁𝐺 (16) -0.085 0.051 -0.047 -0.035 0.029 0.039 -0.089 -0.064 0.019 0.034 -0.020 -0.196 -0.188 0.026 0.169 1.000 -0.083

0.000 0.000 0.000 0.000 0.004 0.000 0.000 0.000 0.061 0.000 0.043 0.000 0.000 0.011 0.000 . 0.000

𝑆𝐷𝐶𝐹𝑂 (17) 0.183 0.146 0.010 -0.075 -0.231 -0.085 -0.223 0.129 -0.036 -0.013 -0.019 0.117 -0.134 -0.274 -0.090 -0.063 1.000

0.000 0.000 0.328 0.000 0.000 0.000 0.000 0.000 0.000 0.206 0.065 0.000 0.000 0.000 0.000 0.000 .

Notes:

Table 3 presents Pearson and Spearman correlations below and above the diagonal respectively.

TABLE 4

Organizational Capital and Cost of Bank Financing – Baseline Analysis

Total Sample Matched Sample

Variables OC IAOC OC IAOC OC IAOC OC IAOC

𝑂𝐶𝐴𝑃𝑇−1 -0.072*** -0.071*** -0.071*** -0.069*** -0.06*** -0.05***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐻𝑂𝐶𝐴𝑃 -0.09*** -0.07***

(0.00) (0.00)

𝑇𝐴𝑁𝐺𝑇−1 -0.058 -0.058 -0.045 -0.045 -0.05 0.16 -0.00 0.10

(0.37) (0.37) (0.37) (0.51) (0.62) (0.15) (0.99) (0.32)

𝐶𝐴𝑃𝑋𝑇−1 -0.077 -0.08 -0.20 -0.40*** -0.24* -0.32**

(0.57) (0.56) (0.16) (0.00) (0.10) (0.02)

𝑅𝑁𝐷𝑇−1 -0.14 -0.13 0.08 0.00 -0.04 0.04

(0.67) (0.69) (0.78) (1.00) (0.85) (0.83)

𝑃𝑃𝑃 -0.38*** -0.38*** -0.38*** -0.38*** -0.39*** -0.35*** -0.38*** -0.36***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑁𝐿𝐷 -0.11*** -0.11*** -0.11*** -0.11*** -0.10*** -0.15*** -0.09*** -0.13***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑁𝑃𝐴𝑅𝑇𝐷 -0.073*** -0.073*** -0.073*** -0.073*** -0.06*** -0.05*** -0.07*** -0.07***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐿𝑃𝐶𝑇 -0.23*** -0.23*** -0.23*** -0.23*** -0.25*** -0.33*** -0.26*** -0.32***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐿𝑅𝐸𝑃 -0.07*** -0.07*** -0.072*** -0.07*** -0.08*** -0.07*** -0.07*** -0.07***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑁𝑈𝑀𝐶𝑂𝑉 0.005*** 0.005*** 0.005*** 0.005*** 0.005*** 0.005*** 0.005*** 0.005***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑇𝐸𝑅𝑀 0.023*** 0.023*** 0.023*** 0.023*** 0.016*** 0.00 0.016*** 0.01*

(0.00) (0.00) (0.00) (0.00) (0.00) (0.80) (0.00) (0.05)

𝑅𝐸𝑉𝐷 -0.34*** -0.34*** -0.34*** -0.34*** -0.33*** -0.34*** -0.32*** -0.34***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑆𝐸𝐶𝑈𝑅 0.01 0.01 0.01 0.01 0.02 0.00 0.02 0.01

(0.53) (0.53) (0.53) (0.53) (0.28) (0.83) (0.24) (0.83)

𝐿𝑁𝐿𝑂𝐴𝑁 -0.036*** -0.036*** -0.036*** -0.036*** -0.04*** -0.04*** -0.04*** -0.05***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑆𝐷𝐶𝐹𝑂 1.11*** 1.11*** 1.15*** 1.15*** 1.32*** 0.66*** 1.03*** 0.62***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐿𝑁𝑇𝐴 -0.17*** -0.17*** -0.17*** -0.17*** -0.18*** -0.14*** -0.17*** -0.14***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑃𝑇𝑅𝑂𝐴 -0.34*** -0.34*** -0.35*** -0.35*** -0.59*** -0.31*** -0.58*** -0.39***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐿𝑉𝑅𝐺 0.56*** 0.56*** 0.56*** 0.56*** 0.64*** 0.64*** 0.64*** 0.57***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑃𝐼𝐹𝑂 -0.98 -0.98 -0.97 -0.97 -0.84*** -0.82*** -0.88*** -0.83***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐴𝑄 1.15*** 1.15*** 1.15*** 1.14*** 1.19*** 0.99*** 1.24*** 0.87***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐶𝐼𝑆 0.72*** 0.72*** 0.72*** 0.72*** 0.71*** 0.75*** 0.71*** 0.73***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑀𝑇𝐵 0.00 0.00 0.00 0.00 -0.00 -0.00 -0.00 -0.00

(0.97) (0.97) (0.97) (0.97) (0.31) (0.61) (0.31) (0.61)

𝐼𝑂𝑊𝑁 -0.19*** -0.19*** -0.19*** -0.19*** -0.20*** -0.31*** -0.22*** -0.29***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐴𝐹𝑂𝐿 -0.02*** -0.02*** -0.016*** -0.016*** -0.02*** -0.02*** -0.02*** -0.02***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝑍𝑆𝐶𝑂𝑅𝐸 -0.018*** -0.018*** -0.018*** -0.018*** -0.01*** -0.01*** -0.01*** -0.01***

(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes Yes Yes Yes Yes

𝑅2 0.61 0.61 0.61 0.61 0.62 0.62 0.62 0.62

𝑂𝑏𝑠 8944 8944 8944 8944 6395 3099 6395 3099

Notes:

Table 4 presents results from the estimation of the equation (3) using both total and propensity score

matched observations. Asterisks above the coefficients represent significance levels where * is used

for p < 10%, ** is used for p < 5% *** is used for p < 1% significance levels. All p-values are

reported in parentheses. Standard errors are clustered by firm and all regressions include industry (2-

digit SIC) effects and time fixed effects. All variables are defined in Appendix A.

TABLE 5

Organizational Capital and Cost of Bank Financing – Control for Corporate Governance

Panel A: Institutional Ownership

Total Sample Matched Sample

HOWN LOWN HOWN LOWN

Variables OC IAOC OC IAOC OC IAOC OC IAOC

𝑂𝐶𝐴𝑃𝑇−1 -0.11*** -0.11*** -0.037* -0.04* -0.063** -0.058* -0.04 -0.04

(0.00) (0.00) (0.09) (0.08) (0.03) (0.03) (0.20) (0.15)

𝑇𝐴𝑁𝐺𝑇−1 -0.085 -0.087 -0.05 -0.05 -0.12 -0.12 -0.03 -0.03

(0.49) (0.49) (0.50) (0.50) (0.36) (0.36) (0.81) (0.81)

𝐶𝐴𝑃𝑋𝑇−1 0.16 0.16 0.067 0.068 0.54 0.54 0.01 0.02

(0.70) (0.70) (0.62) (0.61) (0.18) (0.18) (0.97) (0.97)

𝑅𝑁𝐷𝑇−1 0.38 0.37 -0.18 -0.18 0.34 0.31 0.65 0.67

(0.51) (0.51) (0.61) (0.62) (0.52) (0.55) (0.25) (0.253

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes Yes Yes Yes Yes

𝑅2 0.64 0.64 0.56 0.56 0.61 0.61 0.61 0.61

𝑂𝑏𝑠 4816 4816 4128 4128 2451 2451 2274 2274

Panel B: Analysts Following

Total Sample Matched Sample

HAFOL LAFOL HAFOL LAFOL

Variables OC IAOC OC IAOC OC IAOC OC IAOC

𝑂𝐶𝐴𝑃𝑇−1 -0.11*** -0.10*** -0.03 -0.03 -0.09** -0.085** -0.08* -0.07*

(0.00) (0.00) (0.13) (0.13) (0.01) (0.02) (0.07) (0.07)

𝑇𝐴𝑁𝐺𝑇−1 -0.04 -0.044 -0.04 -0.04 -0.20* -0.20* 0.01 0.01

(0.67) (0.67) (0.67) (0.67) (0.07) (0.07) (0.90) (0.90)

𝐶𝐴𝑃𝑋𝑇−1 -0.13 -0.13 -0.28 -0.28 0.20 0.20 -0.18 -0.17

(0.46) (0.46) (0.09) (0.09) (0.46) (0.45) (0.36) (0.39)

𝑅𝑁𝐷𝑇−1 0.31 0.29 -0.56 -0.56 0.52 0.50 -0.03 -0.06

(0.46) (0.46) (0.16) (0.15) (0.29) (0.31) (0.96) (0.92)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes Yes Yes Yes Yes

𝑅2 0.65 0.65 0.50 0.50 0.60 0.60 0.51 0.51

𝑂𝑏𝑠 4824 4824 4120 4120 2125 2125 2217 2217

Notes:

Table 5 presents results from the estimation of the equation (3) controlling for the level of institutional ownership (Panel A) and the number of analysts following as alternative

measures of corporate governance. The table presents results for both total and propensity score matched samples. Asterisks above the coefficients represent significance levels

where * is used for p < 10%, ** is used for p < 5% *** is used for p < 1% significance levels. All p-values are reported in parentheses. Standard errors are clustered by firm and

all regressions include industry (2-digit SIC) effects and time fixed effects. All variables are defined in Appendix A.

TABLE 6

Organizational Capital and Cost of Bank Financing – Control for State-Level Non-Compete Law

Intensity

Panel A: Total Sample Analysis of Non-Compete Enforcement Index

Variables NCEI>MED NCEI<MED NCEI>MED NCEI<MED

𝑂𝐶𝑇−1 -0.05* -0.06

(0.02) (0.18)

𝐼𝐴𝑂𝐶𝑇−1 -0.05** -0.05

(0.02) (0.19)

𝑇𝐴𝑁𝐺𝑇−1 -0.21 0.09 -0.2111 0.09

(0.03) (0.31) (0.03) (0.32)

𝐶𝐴𝑃𝑋𝑇−1 -0.06 -0.08 -0.05 -0.08

(0.84) (0.58) (0.86) (0.59)

𝑅𝑁𝐷𝑇−1 -0.66 -0.23 -0.65 -0.24

(0.19) (0.61) (0.19) (0.59)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.65 0.61 0.65 0.61

𝑂𝑏𝑠 4516 4371 4516 4371

Panel B: Matched Sample Analysis of Non-Compete Enforcement Index

Variables NCEI>MED NCEI<MED NCEI>MED NCEI<MED

𝑂𝐶𝑇−1 -0.076** -0.049

(0.02) (0.26)

𝐼𝐴𝑂𝐶𝑇−1 -0.073** -0.05

(0.02) (0.26)

𝑇𝐴𝑁𝐺𝑇−1 -0.19 -0.01 -0.20 -0.01

(0.06) (0.95) (0.06) (0.93)

𝐶𝐴𝑃𝑋𝑇−1 0.02 -0.14 0.03 -0.14

(0.94) (0.59) (0.91) (0.60)

𝑅𝑁𝐷𝑇−1 -0.73 -0.47 -0.75 -0.47

(0.15) (0.42) (0.14) (0.41)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.67 0.61 0.67 0.60

𝑂𝑏𝑠 3617 3555 3617 3555

Panel C: Matched Sample Analysis of Non-Compete Enforcement Index using Garmaise (2009) Index

Variables NCEI>MED NCEI<MED NCEI>MED NCEI<MED

𝑂𝐶𝑇−1 -0.11** -0.05

(0.02) (0.27)

𝐼𝐴𝑂𝐶𝑇−1 -0.10** -0.05

(0.03) (0.25)

𝑇𝐴𝑁𝐺𝑇−1 -0.43 -0.05 -0.43 -0.05

(0.00) (0.74) (0.00) (0.73)

𝐶𝐴𝑃𝑋𝑇−1 0.07 -0.55 0.08 -0.54

(0.85) (0.13) (0.81) (0.13)

𝑅𝑁𝐷𝑇−1 -0.36 -0.20 -0.39 -0.19

(0.61) (0.78) (0.59) (0.79)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.69 0.65 0.69 0.65

𝑂𝑏𝑠 1609 1488 1609 1488

Table 6 tests the effects of state-level differences in NCAs on cost of debt capital associated with

organizational capital. Panel A and B presents results for states with Non-Compete Enforceability

Index greater/(less) than the sample median level for total and propensity score matched samples,

respectively. Panel C uses state-level non-compete index calculated by Garmaise (2009) for years

2004 and before. Asterisks above the coefficients represent significance levels where * is used for p <

10%, ** is used for p < 5% *** is used for p < 1% significance levels. All variables are explained in

greater detail in Appendix A.

TABLE 7

Organizational Capital and Non-Price Loan Terms

Total Sample Matched Sample

Variables LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝑇−1 -0.042*** 0.06*** -0.03 0.07***

(0.00) (0.00) (0.15) (0.00)

𝑇𝐴𝑁𝐺𝑇−1 -0.03 -0.09 -0.07 -0.20**

(0.62) (0.22) (0.51) (0.02)

𝐶𝐴𝑃𝑋𝑇−1 0.28* 0.14 0.23 0.50

(0.10) (0.36) (0.50) (0.04)

𝑅𝑁𝐷𝑇−1 0.02 -0.80*** -0.06 -0.89***

(0.95) (0.00) (0.88) (0.00)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.60 0.74 0.60 0.75

𝑂𝑏𝑠 8944 8944 6451 6451

Notes:

Table 7 presents the results for the equation (4) which tests for the link between non-price loan terms

and organizational capital using both total and matched samples. Asterisks above the coefficients

represent significance levels where * is used for p < 10%, ** is used for p < 5% *** is used for p <

1% significance levels. All p-values are reported in parentheses. Standard errors are clustered by firm

and all regressions include industry (2-digit SIC) effects and time fixed effects. All variables are

defined in Appendix A.

TABLE 8

Organizational Capital and Non-Price Loan Terms – Control for Corporate Governance

Panel A: Institutional Ownership

Total Sample Matched Sample

HOWN LOWN HOWN LOWN

Variables LNCOV LNLOAN LNCOV LNLOAN LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝑇−1 -0.03 0.058*** -0.04 0.05*** -0.02 0.04* -0.066* 0.06**

(0.31) (0.00) (0.15) (0.00) (0.57) (0.10) (0.10) (0.05)

𝑇𝐴𝑁𝐺𝑇−1 -0.01 -0.11 -0.04 -0.10 0.13 -0.15 0.02 -0.03

(0.95) (0.34) (0.69) (0.31) (0.46) (0.27) (0.89) (0.85)

𝐶𝐴𝑃𝑋𝑇−1 0.51 0.46 0.30* 0.02 0.26 0.61* 0.23 -0.03

(0.31) (0.13) (0.10) (0.92) (0.67) (0.06) (0.58) (0.94)

𝑅𝑁𝐷𝑇−1 0.24 -0.78 -0.02 -0.77*** 0.37 -0.36 1.45** -0.77

(0.64) (0.03) (0.97) (0.00) (0.65) (0.40) (0.03) (0.11)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes Yes Yes Yes Yes

𝑅2 0.58 0.74 0.61 0.73 0.62 0.70 0.59 0.74

𝑂𝑏𝑠 4816 4816 4128 4128 2451 2451 2274 2274

Panel B: Analysts Following

Total Sample Matched Sample

HAFOL LAFOL HAFOL LAFOL

Variables LNCOV LNLOAN LNCOV LNLOAN LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝐴𝑃𝑇−1 -0.07* 0.045 -0.01 0.055*** -0.01 0.04 -0.12 0.04

(0.09) (0.13) (0.68) (0.00) (0.80) (0.32) (0.17) (0.46)

𝑇𝐴𝑁𝐺𝑇−1 -0.30 0.05 0.02 -0.09 -0.05*** 0.08 -0.06 -0.06

(0.09) (0.71) (0.85) (0.30) (0.00) (0.60) (0.79) (0.79)

𝐶𝐴𝑃𝑋𝑇−1 1.24** -0.37 0.23 0.11 1.22* -0.27 0.17 0.17

(0.04) (0.35) (0.20) (0.53) (0.05) (0.52) (0.82) (0.82)

𝐻𝑀𝐽𝐶 0.01 -0.00 0.13 -0.04 0.08 -0.02 0.04 -0.18

(0.92) (0.97) (0.09) (0.46) (0.21) (0.70) (0.77) (0.02)

𝐻𝑀𝐽𝐶_𝑂𝐶𝐴𝑃𝑇−1 0.02 -0.06 -0.10 0.08 -0.05 -0.06 -0.03 0.07

(0.73) (0.15) (0.17) (0.26) (0.51) (0.23) (0.79) (0.47)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes Yes Yes Yes Yes

𝑅2 0.57 0.78 0.62 0.72 0.58 0.77 0.67 0.75

𝑂𝑏𝑠 3835 3835 5109 5109 2865 2868 1182 1182

Notes:

Table 8 presents the results for the equation (4) which tests for the link between non-price loan terms and organizational capital controlling for the level of

institutional ownership (Panel A) and the number of analysts following as alternative measures of corporate governance. The table presents results for both

total and propensity score matched samples. Asterisks above the coefficients represent significance levels where * is used for p < 10%, ** is used for p < 5%

*** is used for p < 1% significance levels. All p-values are reported in parentheses. Standard errors are clustered by firm and all regressions include industry

(2-digit SIC) effects and time fixed effects. All variables are defined in Appendix A.

TABLE 9

Organizational Capital and Cost of Bank Financing – Control for State-Level Non-Compete Law

Intensity

Panel A: Total Sample Analysis of Non-Compete Enforcement Index

Variables NCEI>M NCEI<MED

LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝐴𝑃𝑇−1 -0.066** 0.046* -0.01 0.06**

(0.01) (0.02) (0.56) (0.02)

𝑇𝐴𝑁𝐺𝑇−1 -0.12 -0.17 0.03 -0.04

(0.35) (0.14) (0.80) (0.70)

𝐶𝐴𝑃𝑋𝑇−1 0.68* -0.078 0.15 0.24

(0.07) (0.81) (0.45) (0.19)

𝑅𝑁𝐷𝑇−1 -0.56 -0.77* 0.26 -0.59**

(0.23) (0.04) (0.56) (0.06)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.61 0.77 0.60 0.74

𝑂𝑏𝑠 4516 4313 4371 4371

Panel B: Matched Sample Analysis of Non-Compete Enforcement Index

Variables NCEI>MED NCEI<MED

LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝐴𝑃𝑇−1 -0.08** -0.01 -0.01 0.06**

(0.04) (0.83) (0.83) (0.01)

𝑇𝐴𝑁𝐺𝑇−1 -0.18 -0.14 -0.14 -0.11

(0.19) (0.23) (0.23) (0.28)

𝐶𝐴𝑃𝑋𝑇−1 0.72* 0.04 0.04 0.26

(0.06) (0.89) (0.89) (0.43)

𝑅𝑁𝐷𝑇−1 -0.25 -0.46 -0.46 -0.83**

(0.62) (0.22) (0.22) (0.04)

𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦 & 𝑇𝑖𝑚𝑒 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.62 0.77 0.77 0.75

𝑂𝑏𝑠 3617 3617 3611 3611

Panel C: Matched Sample Analysis of Non-Compete Enforcement Index using Garmaise (2009) Index

Variables NCEI>MED NCEI<MED

LNCOV LNLOAN LNCOV LNLOAN

𝑂𝐶𝐴𝑃𝑇−1 -0.06 0.01 -0.14** 0.09***

(0.34) (0.77) (0.03) (0.00)

𝑇𝐴𝑁𝐺𝑇−1 -0.25 -0.22 0.02 -0.07

(0.35) (0.15) (0.93) (0.65)

𝐶𝐴𝑃𝑋𝑇−1 1.33* -0.09 0.24 0.33

(0.05) (0.82) (0.75) (0.34)

𝑅𝑁𝐷𝑇−1 -0.00 0.30 1.14 -2.30***

(1.00) (0.60) (0.37) (0.00)

𝐼 & 𝑇 & 𝑆 𝐹𝐸 Yes Yes Yes Yes

𝑅2 0.57 0.80 0.54 0.79

𝑂𝑏𝑠 1609 1609 1488 1488

Table 9 tests the effects of state-level differences in NCAs on non-price loan terms. Panel A and B

presents results for states with Non-Compete Enforceability Index greater/(less) than the sample

median level for total and propensity score matched samples, respectively. Panel C uses state-level

non-compete index calculated by Garmaise (2009) for years 2004 and before. Asterisks above the

coefficients represent significance levels where * is used for p < 10%, ** is used for p < 5% *** is

used for p < 1% significance levels. All variables are explained in greater detail in Appendix A.

APPENDIX A

Variable Definitions

Measuring Organizational

Capital Stock

OC

=Organizational capital measured as in Eisfeldt and Papanikolaou (2013).

𝑂𝐶𝑖,𝑡 = (1 − 𝑑𝑂𝐶) × 𝑂𝐶𝑖,𝑡−1 +𝑆𝐺&𝐴𝑖,𝑡

𝑐𝑝𝑖𝑡

𝑂𝐶𝑖,0 =𝑆𝐺&𝐴𝑖,1

𝑔 + 𝑑𝑂𝐶

The model uses Selling, General and Administrative (Compustat: xsga) to calculate

the organizational capital and offers a proxy level of accumulated organizational

capital should such investments would have been capitalized. 𝑂𝐶𝐴𝑃𝑖,0 is the initial

level of organizational capital for firm i at time 0 which is then used to recursively

estimate firm-level organizational capital stock. As in Eisfeldt and Papanikolaou

(2013) the model assumes 10% growth rate in investment in organizational capital

and 15% depreciation rate. Firm-year level organizational capital is then estimated

by accumulating inflation-adjusted (cpi) SG&A and appropriately depreciating these

investments over time. By default, all missing values of SG&A is treated as zero as

in Eisfeldt and Papanikolaou (2013).

IAOC

=is the industry-adjusted level of organizational capital estimated as the difference

between firm-level organizational capital (OC) and the industry-mean level of

organizational capital calculated using 2-digit SIC codes.

HOCAP

=is the dummy indicator that takes the value of 1/(0) if firm-level/industry-adjusted

organizational capital is above/(below) sample median levels.

OCAP =refers to firm-level and industry adjusted forms of organizational capital.

Other forms of

Investments

TANG

=is the asset tangibility measure using net property, plant and equipment that is

adjusted for depreciation deduction (Compustat: PPENT) scaled by total assets.

CAPX

=is the fixed capital investment measured using CAPX (Compustat: CAPX) scaled

by total assets.

RND

=is research and development expenses measured using RND (Compustat: RND)

scaled by total assets.

NCA-Intensity

NCEI

=is the measure of state level NCA intensity. The Non-Compete Enforceability

Index is based on the twelve question that Malsberger (2013) set forth that examines

jurisdictional enforcement intensity of non-compete laws. Next, I personally track all the legislation changes in state-level non-compete agreements (NCAs), manually

read each “passed” bill when necessary and update NCEI for the years following

2013 to 2015. I also utilize the index estimated in Garmaise (2009) by limiting the

data for periods between 1994 and 2004. The index can range from 0, for states that

restricts non-compete agreements and a possible 12 for states that has the strictest

applications of such clauses

Loan Specific Variables

SPREAD = Loan spread required by banks obtained from Thomson Deals database.

LNSPREAD = Natural log of the SPREAD

NLD

= Dummy indicator which takes the value 1/(0) if the loan facility has greater/(less)

than median number of lead arrangers and controls for syndicate-lead-level risk

diversification

PPP = Dummy indicator for loans that include performance pricing provisions.

COV = The number of covenants included in a loan.

LPCT

= The proportion of loan held by the syndicate arrangers. Unlike the past research

(e.g., Sufi, 2007), the measure aims to capture the total portion of loan held by the

lead agents altogether. Therefore, if four lead arrangers hold half of the total loan

amount altogether that is the ratio I use in LEADPCT and not 12.5 percent (50%/4)

for each lead bank.

LREP

= Lead arranger reputation. I classify top five syndicate arrangers per given year in

Thomson Deals database as the most reputable lenders. Next, I identify loans with

the number of reputable lenders in the top quartile of the total sample distribution

(LREP).

TERM =Average loan maturity.

LNLOAN =Natural logarithm of the outstanding loan amount

REVD

=Dummy indicator that takes the value of 1 if the loan is a revolving credit facility

and 0 otherwise.

SECUR

= Dummy indicator that takes the value of 1 if the loan is secured via collateral and

0 otherwise.

Governance Variables

OWN

=Percentage of institutional ownership obtained from Thomson Institutional

Holdings database.

AFOL =The number of analysts following the firm. Obtained from IBES summary files.

Other Control Variables

PTROA =Total Pre-tax Income (Compustat: PI) divided by total assets (Compustat: AT).

LNTA =Natural logarithm of total assets.

PIFO =Pre-tax income from foreign operations (Compustat: PIFO) divided by total assets.

LVRG =Long-term debt (Compustat: DLTT) divided by total assets.

PPE =Net property, plant and equipment (Compustat: PPENT) scaled by total assets.

CIS

=Four-quarter moving average of the spread of commercial and industrial loan rates

(loans worth more than $1MN) over the federal fund rates.

CFNAI =A measure of overall economic activity using Chicago Fed National Activity Index

AQ

=Following Cook et al. (2015), accrual quality is calculated as the standard

deviation of the firm-level residuals as in Francis et al. (2005) from the following

model.

𝑇𝐶𝐴 = 𝐶𝐹𝑂𝑡−1 + 𝐶𝐹𝑂𝑡 + 𝐶𝐹𝑂𝑡+1 + ∆𝑆𝐴𝐿𝐸 + 𝑃𝑃𝐸𝐺𝑇 + 𝜀

In the above model total current accruals TCA is estimated as [Compustat: ∆𝐴𝐶𝑇 −

∆𝐿𝐶𝑇 − ∆𝐶𝐻𝐸 + ∆𝐷𝐿𝐶]. CFO is income before extraordinary items (Compustat:

IB) minus total current accruals minus depreciation and amortization (Compustat:

DP). All variables are scaled by total assets. The model is estimated for each 2 digit

SIC code with 15 or more observations.

SDCFO = is the standard deviation of cash flows from operations (Compustat: OANCF)

ZSCORE = is the measure of financial health as in Altman (1968).

MTB

= The ratio of market value of equity [Compustat: 𝑃𝑅𝐶𝐶_𝐹 × 𝐶𝑆𝐻𝑂] to book value

of equity [Compustat: CEQ].

APPENDIX B

Propensity Score Matching

Univariate Tests of Observable Firm Characteristics for PSM Pairs

Panel A: Treatment Variable – HOCAP

Variables Treatment Control diff p-value

𝑇𝐴𝑁𝐺 0.279 0.27 0.009 0.79

𝑅𝑁𝐷 0.0175 0.018 -0.005 0.15

𝐶𝐴𝑃𝑋 0.056 0.055 0.001 0.38

𝑃𝑇𝑅𝑂𝐴 0.053 0.052 0.001 0.82

𝐿𝑉𝑅𝐺 0.253 0.254 -0.001 0.85

𝐿𝑁𝑇𝐴 7.13 7.18 -0.005 0.18

𝑍𝑆𝐶𝑂𝑅𝐸 3.58 3.49 0.09 0.25

Panel B: Treatment Variable – NCEI

Variables Treatment Control diff p-value

𝑇𝐴𝑁𝐺 0.31 0.30 0.01 0.05*

𝑅𝑁𝐷 0.02 0.018 0.02 0.13

𝐶𝐴𝑃𝑋 0.062 0.062 0.00 0.87

𝑃𝑇𝑅𝑂𝐴 0.051 0.051 0.00 0.95

𝐿𝑉𝑅𝐺 0.256 0.26 -0.007 0.15

𝐿𝑁𝑇𝐴 7.25 7.18 0.07 0.05*

𝑍𝑆𝐶𝑂𝑅𝐸 3.40 3.40 0.00 0.95

Panel C: Treatment Variable – OWN

Variables Treatment Control diff p-value

𝑇𝐴𝑁𝐺 0.31 0.30 0.01 0.50

𝑅𝑁𝐷 0.018 0.016 0.002 0.16

𝐶𝐴𝑃𝑋 0.058 0.058 0.00 0.99

𝑃𝑇𝑅𝑂𝐴 0.06 0.06 0.00 0.68

𝐿𝑉𝑅𝐺 0.26 0.25 0.01 0.15

𝐿𝑁𝑇𝐴 7.07 7.13 -0.06 0.12

𝑍𝑆𝐶𝑂𝑅𝐸 3.39 3.45 -0.06 0.53

Panel D: Treatment Variable – AFOL

Variables Treatment Control diff p-value

𝑇𝐴𝑁𝐺 0.31 0.32 -0.01 0.72

𝑅𝑁𝐷 0.017 0.016 0.001 0.20

𝐶𝐴𝑃𝑋 0.069 0.064 0.005 0.02**

𝑃𝑇𝑅𝑂𝐴 0.05 0.045 0.005 0.31

𝐿𝑉𝑅𝐺 0.28 0.27 0.01 0.32

𝐿𝑁𝑇𝐴 6.97 7.06 -0.09 0.00***

𝑍𝑆𝐶𝑂𝑅𝐸 3.49 3.43 0.06 0.47

Notes:

The table above presents univariate tests of observable firm characteristics for propensity score

matched pairs. Panel A and Panel B use above/below median levels of organizational capital

(HOCAP) and Non-Compete Enforcement Intensity Index (NCEI) and Panel C and Panel D use

above/below median levels of institutional ownership and analyst coverage as treatment variables in

related probit regressions of matching procedure, respectively. Asterisks above the coefficients

represent significance levels where * is used for p < 10%, ** is used for p < 5% *** is used for p <

1% significance levels. All variables are explained in greater detail in Appendix A.