Development costs capitalization and debt financing

• Author: Ioannis Tsalavoutas,Brigitte Eierle,Andreas Kreß
• Date: 01 May 2019
• DOI: http://doi.org/10.1111/jbfa.12370
• Published date: 01 May 2019

DOI: 10.1111/jbfa.12370

Development costs capitalization and debt

financing

Andreas Kreß1Brigitte Eierle1Ioannis Tsalavoutas2

1University of Bamberg, Germany

2University of Glasgow, Adam Smith Business

School, Scotland, UK

Correspondence

AndreasKreß, University of Bamberg, Room

03.38,Feldkirchenstraße 21, 96047 Bamberg,

Germany.

Email:andreas.kress@uni-bamberg.de

Abstract

This study investigates debt market effects of research and devel-

opment (R&D) costs capitalization, using a global sample of public

bonds and private syndicated loans issued by public non-financial

firms. Firstly, we show that firms capitalize larger amounts of R&D

in a year when they exhibit a propensity for issuing bonds, rather

than borrowing funds privately from the syndicated loan market, in

the subsequent year.Secondly, we provide evidence that capitalized

R&D investments reduce the cost of debt. We infer that debt mar-

ket participants are able to identify firms’ motives for R&D capital-

ization, as we find a reduction in the cost of debt only for those firms

that do not show indications of employing R&D capitalization for

earnings management reasons. Indeed, only for this sub-sample of

firms, the amount of capitalized R&D contributes positively to future

earnings. Weconfirm that R&D capitalization is positively associated

with audit fees and thus can be deemed to be a signaling device.

Lastly, we find that it is the amount of R&D a firm is expected to

capitalize and not the discretionary counterparts, which facilitates

afirm's access to public debt markets, reduces bond and syndicated

loan prices, and contributes to future benefits.

KEYWORDS

cost of debt, debt financing, debt markets, R&D capitalization, R&D,

research and development

1INTRODUCTION

International Accounting Standard (IAS) 38 Intangible Assets, within International Financial Reporting Standards

(IFRS), separates development expenditures from research costs and requires different treatment for each compo-

nent. Research outlays must be expensed as incurred, but development costs must be capitalized when certain con-

ditions are met. This contrasts with the requirements of United States Generally Accepted Accounting Principles (US

GAAP) where both research costs and development expenditures must be expensed. Developmentcosts capitalized

under IAS 38 relate to R&D projects which are closer to being used or sold and hence bear less uncertainty regarding

their future outcome. However,as fulfillment of the restrictive recognition criteria under IAS 38 requires managers to

636 c

2019 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2019;46:636–685.

KREß ET AL.637

use proprietary information and exercise subjective judgement, R&D capitalization under IFRS is open to managerial

discretion.

We first investigate whether or not managers use this discretion as a means of facilitating borrowing from pub-

lic debt markets, rather than entering private debt markets (i.e.,the syndicated loan market). Thus, we examine if the

amount of R&D a firm capitalizes in a given year is associated with a firm's choice of the source of debt financing for

the subsequent year.We then analyze the value relevance of R&D capitalization for debt markets by investigating the

effect of capitalized development costs in a given year on the cost of public debt and private debt for the subsequent

year.Tothe best of our knowledge, evidence on these two issues is absent. Prior studies examining the debt market con-

sequences of R&D investments focus exclusivelyon US firms (Ciftci & Darrough, 2016; Eberhart, Maxwell, & Siddique,

2008; Shi, 2003) and thus rely on a setting for which discretionary R&D capitalization cannot be observed.

We first hypothesize that firms capitalize larger amounts of R&D in a year when they show a higher propensity

subsequently to raise funds from the public debt market ratherthan the syndicated loan market (H1). The public debt

market and syndicated loan marketare regarded as mutual substitutes (e.g. Altunba ¸s,Kara, & Marques-Ibanez, 2010),

because in both markets firms can raise considerable amounts of funds with comparable maturity terms. However,

the markets differ strongly in information asymmetry (Altunba¸s et al., 2010; Bharath, Sunder, & Sunder,2008). Firms

accessing the syndicated loan market are able to communicate the future success of their R&D projects via private

channels, while bondholders in the public debt marketdo not have access to such channels (Bharath et al., 2008; Florou

& Kosi, 2015). Thus, information in publicly available financial statements can be more important for bondholders in

this context (Bharath et al., 2008; Gorton & Winton, 2003; Marshall, Mccann, & Mccolgan, 2016). Consequently, for

our chosen context, firms that plan to access public debt markets havegreater incentives to signal the future success

of their R&D investments and are willing to incur the necessary costs – including higher audit fees – to capitalize the

corresponding development expenditures.

Further,we hypothesize a negative association between capitalized R&D and the cost of public (H2) and private debt

(H3). These hypotheses are motivated by the asymmetric payoffstructure of debtholders. They bear the full extent of

the downside risk, while their return is restricted to a fixed interest rate.Thus, debtholders are more concerned about

bad news, which may affect their downside risk (Ball, Bushman, & Vasvari, 2008; Easton, Monahan, & Vasvari,2009).

As innovation projects show a high degree of technical and commercial uncertainty as well as a low success rate (Lev,

2001), the risk component and thus the trade-off between future benefits and risk of R&D investmentsis more preva-

lent for debt markets (Ciftci& Darrough, 2016; Shi, 2003). Capitalized development costs should, therefore, be of par-

ticular importance to debt marketparticipants. Arguably, the high reporting costs and audit effort involved in recording

capitalized development costs (Cheng, Lu, & Kuo, 2016; De George, Ferguson,& Spear, 2013; Kuo & Lee, 2017) could

further assist debtholders to perceive capitalized development costs as a signal of genuine future economic benefits

resulting from less risky R&D projects. Hence, capitalized R&D investments should be priced positively, resulting in

reduced debt costs.

We test these hypotheses by using a global sample of bonds and private syndicated loans issued by public, non-

financial firms in countries which mandated IFRS or fully converged their local GAAP with IFRSfrom 2005 onwards.

Consistent with our first hypothesis, the propensity to borrow in the public debt marketrather than the private syn-

dicated loan market in a givenyear is positively associated with higher amounts of R&D having been capitalized during

theprevious year. Regarding the cost of public debt, we find that R&D costs are priced differently for firms that expense

all of their R&D investments (Expensers) and for firms that capitalize all or some of their R&D investments(Capitaliz-

ers). For Expensers, R&D expenditure reduces bond prices. This is in line with the literature on US firms (Eberhart

et al., 2008), where firms expense all their R&D investments. For Capitalizers, the capitalized and expensedportions

of R&D investments are regarded differently by bond investors. While the amount of R&D capitalized during a year

reduces the cost of public debt, the expensedcomponent is not priced. Regarding the effect of R&D investments on the

cost of private debt, we find that R&D costs of Expensers are not priced in syndicated loan deals. However, in line

with our hypothesis, we find that the amount of capitalized R&D reduces syndicated loan prices for Capitalizers,

whereas our results show no significant effect of the expensed component. These findings provide evidencethat debt

investors overallregard the capitalized part of R&D investments as a signal of reduced risk from R&D projects. This is

638 KREß ET AL.

consistent with the restrictive conditions in IAS 38, which are directed towards indications of likely success of firms’

capitalized development costs.

In further analysis, we show that debt marketparticipants are able to identify firms’ motives for R&D capitalization,

as capitalized development costs are priced only for firms whose capitalized amount is not attributed to earnings man-

agement incentives. Furthermore, for capitalizing firms we find that only the capitalizedamount contributes to future

earnings, while the expensed counterpart is not related to future benefits. In fact, the positive association between

capitalized R&D and firms’ future earnings holds only for firms that do not show indications of employing R&D capital-

ization for earnings management reasons. Additionally,we document a significant positive relationship between R&D

capitalization and a firm's audit fees, implying that R&D capitalization as required by IAS 38 is a costly activity and so

can be deemed to be a signaling device. Lastly, when we separatethe amount of R&D that a firm capitalized during a

year into expected and discretionary components, we show that it is the amount of R&D a firm is expectedto capital-

ize, which facilitates a firm's access to public debt markets, reduces bond and syndicated loan prices, and contributes

to future benefits.

We contribute to the literature firstly in adressing the call of Christensen, Nikolaev, and Wittenberg-Moerman

(2016, p. 427) for future research by providingevidence of how ‘the choice among different accounting methods’ facil-

itates debt contracting ‘when economic incentives to deliver an informative measure of the economic performance

are present’.Bharath et al. (2008), Dhaliwal, Khurana, and Pereira (2011), Florou and Kosi (2015) and Ball, Hail, and

Vasvari (2017) show that firms with lower information asymmetry exhibita greater propensity for raising funds from

public debt marketsthan private debt markets. Thus, our analysis indicates how a specific accounting treatment, which

is at the heart of the accounting choice literature, is used to reduce information asymmetries to bondholders. Sec-

ondly, our study adds to recent literatureby identifying R&D capitalization as a further mechanism that helps ‘to cor-

rect the potential misvaluation of firms’ R&D investments’ (Zhang & Toffanin,2018, p. 25). Our findings which demon-

strate a positive association between development costs capitalized and future earnings as well as audit fees imply

that R&D capitalization under IAS 38 is a costly,albeit effective, signaling device for managers to convey private infor-

mation about the future success of their R&D projects to debtholders. Thirdly, our investigations respond to the call

for research on the value relevance of accounting numbers to debtholders (Givoly,Hayn, & Katz, 2017; Holthausen &

Watts, 2001). Directly relevant to our research, Givoly et al. (2017, p. 69) state that ‘whether capitalization [ofintan-

gible assets] is beneficial to creditors is an open question’.Ours is the first study to show that capitalized development

costs reduce the costs of both public and private debt.

Beyondthe academic contributions stemming from our analyses, the findings will present valuable information both

for standard setters and regulators. Forexample, in 2018 the UK's Financial Reporting Council (FRC) initiated a project

not only to review current requirements and practice for the business reporting of intangibles, but also subsequently

to develop practical proposals for their improvement.1Inthe feedback statement of its research agenda consultation

the European Financial Reporting Advisory Group (EFRAG) is also proposing research on this area in the near future.2

The remainder of this paper is structured as follows. Section 2 discusses the relevant literature and hypotheses

development. Section 3 describes the sample selection process and research design. Section 4 provides descriptive

and multivariate analysis results. Section 5 presents sensitivity tests and Section 6 concludes.

2LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT

2.1 Debate about the accounting treatment of R&D

Lev (2001) argues that R&D investments may enable firms to obtain a temporary monopoly in the market, allow-

ing them to extract substantial future cash flows. Consistent with this, prior literature provides evidence that

1See:https://www.frc.org.uk/accountants/accounting-and-reporting-policy/research/intangibles-how-can-business-reporting-do-better.

2See:https://www.efrag.org/News/Project-324/Feedback-Statement–2018-EFRAG-Research-Agenda-Consultation.