Sticky behavior of R&D costs and its relevance to firm value: Evidence from listed companies in Korea

Kym, Moon Hyun Article Sticky behavior of R&D costs and its relevance to firm value: Evidence from listed companies in Korea Global Business & Finance Review (GBFR) Provided in Cooperation with: People & Global Business Association (P&GBA), Seoul Suggested Citation: Kym, Moon Hyun (2023) : Sticky behavior of R&D costs and its relevance to firm value: Evidence from listed companies in Korea, Global Business & Finance Review (GBFR), ISSN 2384-1648, People & Global Business Association (P&GBA), Seoul, Vol. 28, Iss. 4, pp. 77-89, https://doi.org/10.17549/gbfr.2023.28.4.77 This Version is available at: https://hdl.handle.net/10419/305909 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/4.0/ I. Introduction This study examines the impact of the sticky behavior of research and development (R&D) costs on firm value using data from listed Korean firms. In R&D activities, costs are recognized as either expenses or assets in accounting. The adoption of Received: Mar. 31, 2023; Revised: May. 19, 2023; Accepted: Jun. 2, 2023 † Moon Hyun Kym E-mail: [email protected] the International Financial Reporting Standards (IFRS) in Korea, which was implemented for all listed companies in 2011, strengthened the capitalization requirements for R&D costs. Therefore, this paper distinguishes between expensed and capitalized R&D costs to measure R&D cost stickiness. The study aims to determine whether both expensed R&D costs and capitalized R&D costs increase the value relevance of earnings and net assets, respectively. Previous studies have shown that costs exhibit asymmetric responses to increases or decreases in GLOBAL BUSINESS & FINANCE REVIEW, Volume. 28 Issue. 4 (AUGUST 2023), 77-89 pISSN 1088-6931 / eISSN 2384-1648∣Https://doi.org/10.17549/gbfr.2023.28.4.77 ⓒ 2023 People and Global Business Association GLOBAL BUSINESS & FINANCE REVIEW www.gbfrjournal.org1) Sticky behavior of R&D Costs and Its Relevance to Firm Value: Evidence from Listed Companies in Korea Moon Hyun Kym† P rofessor, Division of Global Business & Technology, HanKuk University of Foreign Studies, Korea A B S T R A C T Purpose: The purpose of this study is to investigate the impact of sticky behavior in R&D costs on firm value. The study aims to gain insights into whether the stickiness of R&D costs is a rational choice based on managerial foresight or if it stems from managerial self-interest. Design/methodology/approach: For the regression analysis, a modified version of Ohlson's accounting-based valuation model (1995) was used to assess the impact of R&D costs on firm value. and the stickiness of R&D costs was measured using the model proposed by Homburg & Nasev (2009). International Financial Reporting Standards were adopted in Korea in 2011, and empirical analysis was subsequently conducted using data from listed companies. Findings: The empirical results were as follows. Firstly, R&D costs reported as expenses showed a negative (-) correlation with firm value, while the stickiness of R&D costs was found to have a positive (+) impact on the value relevance of earnings. Secondly, R&D costs reported as assets did not have a significant impact on firm value, and the stickiness of R&D costs did not have a significant impact on the value relevance of net assets. Research limitations/implications: Overall, the findings of this study support the accounting treatment of R&D costs as expenses. Moreover, the study reveals that the sticky behavior of R&D costs helps alleviate the negative value relevance of R&D expenses. Originality/value: This study distinguished between capitalized R&D costs and expensed R&D costs, and examined their value relevance by investigating the impact of sticky behavior on earnings and net assets. Keywords: Expensed R&D costs, Capitalized R&D costs, Sticky Cost Behavior, Value Relevance ⓒ Copyright: The Author(s). This is an Open Access journal distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution , and reproduction in any medium, provided the original work is properly cited. GLOBAL BUSINESS & FINANCE REVIEW, Volume. 28 Issue. 4 (AUGUST 2023), 77-89 78 sales (Anderson et al., 2003; Banker and Chen, 2006; Lee, 2012; Park and Sunwoo, 2016). Since Anderson et al. (2003), the asymmetry of cost behavior has been referred to as sticky cost behavior. It occurs when the rate of increase in costs due to an increase in sales is less than the rate of decrease in costs due to a decrease in sales. Subsequent studies have confirmed the existence of cost stickiness in specific industries (Balakrishnan et al., 2004; Subramaniam and Watson, 2016; Cheung et al., 2016). These findings imply that the sticky behavior of costs stems from specific cost elements. This study focuses on the behavior of R&D costs as a cost element. Investment in R&D activities is critical for a company's growth and development, requiring continuous resource allocation. R&D costs are expected to be sticky because they need to be maintained consistently, even during periods of reduced sales. The process of allocating resources over time results in the incurrence of capital adjustment costs. If the decrease in sales is temporary, it may be rational to bear the resource adjustment costs temporarily, and as a result, the cost will sticky behavior (Balakrishnan et al., 2004; Chen et al., 2012; Kitching et al., 2016). On the other hand, managers have an incentive to maintain the resources they can control. Despite a decrease in sales, a manager may seek to increase their own personal utility by maintaining the size of the business without disposing of assets or reducing employees. Overinvestment in resources or the manager's privileged consumption can cause sticky behavior in costs (Anderson et al. 2003; Chen et al. 2012). The relationship between earnings and stock returns can explain why costs exhibit sticky behavior, leading to an asymmetric relation between earnings and stock returns (Banker et al., 2016; Ha et al., 2021). This study extends research on the impact of cost stickiness on stock returns to the impact on stock prices. If the sticky behavior of R&D costs is based on rational choices made by managers with regard to future sales prospects and adjustment costs, it will increase the value relevance of accounting numbers. However, if the sticky behavior of R&D costs is driven by managers' private choices to maintain control over resources, it will not increase the value relevance of accounting numbers. Considering the duality of investment in R&D costs as both expenses and assets, accounting standards require the separation of R&D costs into those that are expensed and those that are capitalized. Research costs and recurring development costs are recognized as expenses, which reduce reported earnings, while non-recurring development costs that are expected to generate future economic benefits are recognized as intangible assets. Previous studies on the impact of R&D costs on firm value do not differentiate between expensed and capitalized R&D expenses (Jo and Lee, 2019; Kim et al., 2021; Park et al., 2021). This study differentiates between R&D costs reported as expenses and R&D costs capitalized as intangible assets, and tests the value relevance of the sticky behavior of these two categories of R&D costs using data from listed Korean firms. The findings of this study provide overall support for considering R&D costs as expenses in accounting. Expensed R&D costs exhibited a negative value relevance to firm value, and the sticky behavior of expensed R&D costs was found to alleviate the negative value relevance. On the other hand, capitalized R&D costs did not demonstrate any value relevance. II. Literature Review and Hypothesis A. Literature Review The traditional cost behavior model is based on the fundamental assumption that the relationship between costs and activity or volume is symmetrical in both its increases and decreases. It is hypothesized that when activity increases (decreases), costs will increase (decrease) in proportion (Noreen, 1991). Anderson et al. (2003) suggested that in the case of increasing sales, the increase in selling and Moon Hyun Kym 79 administrative expenses is smaller than the decrease in these expenses in the event of decreasing sales, resulting in cost stickiness. The relative reduction in cost may not be significant during periods of declining sales (Balakrishnan et al., 2004; Chen et al., 2012; Kitching et al., 2016). Research studies focusing on specific industries have demonstrated that costs in the IT and pharmaceutical/biotechnology industries exhibit sticky cost behavior (Balakrishnan et al.,2004; Cannon, 2014; Kim, 2019; Moon et al, 2020; Hwang and Park, 2020). Cannon (2014) found that the US airline industry exhibits cost sticky behavior, with airlines attempting to lower prices to utilize existing capacity when demand decreases, and trying to increase capacity when demand increases. These findings suggest that there may be differences in sticky behavior among cost elements and that specific industries or costs may exhibit a sticky behavior due to high adjustment costs. Studies conducted in Korea after the adoption of IFRS in 2011 have also reported similar results. When a decrease in sales is temporary and an increase in sales is expected, costs showed a sticky behavior (Lee, 2012; Park and Sunwoo, 2016). Kim (2019) demonstrated that R&D costs of IT companies were more rigidly downward compared to non-IT companies in an analysis of listed companies between 2002 and 2017. Moon et al, (2020) confirmed the sticky behavior of R&D costs in listed digital companies between 2010 and 2017. Yoon (2021) analyzed technology companies in 41 countries from 2010 to 2019 and demonstrated that R&D costs did not decrease as much as sales, even when sales decreased. Cost stickiness leads to an asymmetric relation between earnings and stock returns. As Banker et al. (2016), this relation is more significant during periods of decreasing sales. A study is needed to verify the relationship between stock prices and cost stickiness. The impact of cost sticky behavior on firm value can be found in the reason why costs are sticky. Overall, the literature suggests that cost stickiness can be explained by adjustment costs and agency problems. Adjustment costs incurred in the process of adjusting resource investments (Banker and Byzalov, 2014). Sticky costs occur because managers deliberately adjust the resources committed to activities (Anderson et al., 2003). It means that the relative reduction in cost may not be significant during periods of declining sales (Balakrishnan et al., 2004; Chen et al., 2012; Banker and Byzalov, 2014; Kitching et al., 2016). If it is expected that sales will increase again after a decrease, it would be inefficient to reduce resource investments and incur adjustment costs because doing so would be counterproductive. Lee (2012) interpreted the increase in costs during periods of decreasing sales as a good signal that future profits will increase. Park et al. (2012) confirmed the meaning of such signals through the reaction of the stock market. Park and Sunwoo (2016) showed that when the manager's forecast is considered reliable, sticky costs are associated with the manager's optimistic outlook on the company's future. From the agency theory perspective, the sticky behavior of costs is explained by the fact that managers pursue their own interests (Chen et al., 2012; Namitha and Shijin, 2016). During periods of declining sales, sticky behavior of costs can decrease reported profits, which creates the possibility of earnings management. Goo (2011) suggested that earnings management such as loss avoidance, earnings smoothing, and Big-Bath are related to sticky cost behavior. Song (2013) showed that companies are reducing selling and administrative expenses in the fourth quarter to avoid reporting losses or to increase reported earnings. In a study by Hwang et al. (2017) analyzing listed companies from 2008 to 2015, it was found that for small-profit companies, labor costs decreased more than sales when sales decreased, indicating that they adjusted earnings by reducing labor costs. Tulcanaza-Prieto and Lee (2022) showed real earnings management (REM) negatively affect firm value by introducing quarterly financial data in the Korean market. The introduction of K-IFRS has strengthened the criteria for recognizing R&D costs as assets, compared to previous standards. Kim et al. (2021) analyzed the trend of value relevance of R&D costs using GLOBAL BUSINESS & FINANCE REVIEW, Volume. 28 Issue. 4 (AUGUST 2023), 77-89 80 listed companies from 2000 to 2014 and found that expensed R&D costs had a stronger positive relationship with profitability over five years. Park et al. (2021) found that only capitalized R&D costs were value relevant in the growth stage, while both expensed and capitalized R&D costs were value relevant in the mature stage. Ji (2018) demonstrated that the higher the reliability of intangible assets, the greater their value relevance. On the other hand, Koo and Kim (2023) showed that ESG ratings have an incremental effect on the value relevance of R&D costs. However, due to arbitrary accounting of R&D costs, the value relevance of R&D information is likely to decrease. As issues arose regarding the accounting of R&D costs in pharmaceutical and biotech companies, the Financial Supervisory Service (FSS) announced supervisory guidelines in 2018. The guidelines required stricter accounting of the capitalization of R&D costs and implemented thematic audits to assess the appropriateness of the accounting of R&D costs in pharmaceutical and biotech companies. Hwang and Park (2020) demonstrated that the accounting of R&D costs is related to deficit avoidance, as there is a higher tendency to capitalize these costs when there are incentives for deficit avoidance compared to when there are not. Noh and Choi (2021) showed that there is a decrease in the tendency to capitalize R&D costs and an increase in their value relevance after theme-based review. Kim and Park (2019) examined 75 pharmaceutical and biotechnology firms and found that most of them expensed their R&D costs, while companies that capitalized these costs provided insufficient disclosures in their footnotes about the type and stage of drug development. Ha et al, (2021) investigated the impact of regulatory guidance on firm value using excess returns as the measure. They found that firms affected by the guidance had lower excess returns compared to those that were not affected. Studying the impact of R&D expenditure on firm value is a pertinent research topic that can shed light on the accounting issues related to R&D costs in Korea. B. Hypothesis According to accounting standards, R&D costs are classified as either expensed or capitalized based on the existence of future economic benefits. Expensed R&D costs are recognized as cost of goods sold or selling, general, and administrative expenses, which decrease reported earnings, while capitalized R&D costs are recognized as intangible assets, which increase reported net assets. A difference in value relevance is expected between expensed R&D costs and capitalized R&D costs. Expensed R&D costs have negative (-) value relevance because they decrease reported earnings, while capitalized R&D costs have positive (+) value relevance because they increase net assets. Since the adoption of IFRS to all listed companies in 2011 in Korea, the criteria for capitalizing R&D costs have been strengthened. The stricter criteria for capitalization may have increased the value relevance of capitalized R&D costs. However, according to Kim et al. (2021), the trend of capitalizing R&D costs has decreased and its persistence has not shown a clear trend. Due to the uncertainty of future economic benefits, the value relevance of R&D costs may not be significant. If the value relevance of R&D costs is significantly negative, recognizing R&D costs as intangible assets would not be different from recognizing them as expenses. Hypothesis 1: Expensed R&D costs have a negative (-) value relevance, while capitalized R&D costs have a positive (+) value relevance. Depending on the alternative perspectives that explain the sticky behavior of costs, there may be differences in the relevance to value. If the sticky behavior of R&D costs is derived from the manager's personal interests, it will be negative for value, while if it is derived from the manager's rational choice regarding future prospects, it will be positive for value. R&D costs are essential and sustainable investments necessary for a company's growth and development, and therefore, adjustment costs for R&D are expected to be higher compared to other cost factors. According Moon Hyun Kym 81 to Banker and Byzalov (2014), costs exhibit a sticky behavior due to the adjustment costs incurred in reallocating resources across periods. It is rational to maintain sticky costs in preparation for future sales growth even if current sales decrease. In this case, sticky costs will exhibit a positive (+) value relevance. On the other hand, managers may choose to maintain discretionary spending during times of decreased sales by keeping idle resources within controllable limits (Chen et al., 2012). Reducing assets or employees relatively less during periods of declining sales leads to a sticky cost behavior. If the stickiness of costs is related to the manager's personal interests, then it will exhibit negative (-) value relevance. The sticky behavior of R&D expenses is expected to affect the value relevance of earnings, while the sticky behavior of R&D assets is expected to affect the value relevance of net assets. R&D costs affect both earnings and net assets, so the sticky behavior of R&D costs is expected to impact firm value through their effects on earnings and net assets. Hypothesis 2: The sticky behavior of expensed R&D costs has a positive (+) effect on the value relevance of earnings, while the stickiness of capitalized R&D costs has a positive (+) effect on the value relevance of net assets. III. Research Design A. Model Specification In this study, Ohlson's valuation model (1995) was used to examine the impact of the stickiness of R&D costs on firm value. Assuming that excess earnings disappear at a constant rate over time, firm value is expressed as a linear combination of bottom-line information, including earnings, book value, and other information. This linear model is commonly used in research on the value relevance of accounting information, as it explains firm value using financial statement numbers. Boonlert-U-Thaia and Duangploy (2015) used Ohlson's (1995) model to verify the relative value relevance of U.S. GAAP and IFRS. Ji (2017) employed the model to validate the effects of IFRS adoption, while Koo and Kim (2023) utilized the model to examine the joint effects of ESG rating and R&D costs.  where; P: stock price E: net income B: net asset Others: other information During the accounting period, R&D costs are reported as expenses such as research expenses and recurring development expenses, and some are reported as intangible assets such as non-recurring development costs. If we distinguish the R&D expenses expensed in net income and capitalized as intangible assets in net assets, the above model can be expressed as follows.      where; RNDE: expensed R&D costs RNDA: capitalized R&D costs The regression model was modified to verify the value relevance of R&D costs. The firm value, represented by the stock price, is a function of net income and net assets, excluding R&D costs, as well as expensed R&D costs (RNDE), capitalized R&D costs (RNDA), and other information. Financial statement variables are measured on a per-share basis. Other information includes firm-specific variables, such as total assets, debt ratio, operating profit margin, sales growth rate, and foreign ownership ratio, based on prior research.       GLOBAL BUSINESS & FINANCE REVIEW, Volume. 28 Issue. 4 (AUGUST 2023), 77-89 82 where; E-RNDE: net income excluding expensed R&D costs RNDE: expensed R&D costs B-RNDA: net asset excluding capitalized R&D costs RNDA: net income excluding AST: ln(asset) DR: debt/owner's equity ROI: operating earnings/total asset GRW: sales growth ratio FR: Foreign ownership percentage The stickiness of R&D costs, recognized as both expenses and assets, may impact the value relevance of both net income and net assets. The stickiness of expensed R&D costs is expected to affect stock prices through net income, while the stickiness of capitalized R&D costs is expected to impact stock prices through net assets. A regression model that accounts for these relationships as interaction terms is as follows.             where;   : stickiness of expensed R&D costs  : stickiness of capitalized R&D costs In the model,  represents the effect of stickiness of expensed R&D costs on the value relevance of net income, and  represents the effect of stickiness of capitalized R&D costs on the value relevance of net assets. If capitalizing or expensing R&D expenses increases the value relevance of earnings and net assets, then  and  are expected to be positive. The firm characteristic variables such as total assets, debt ratio, operating profit margin, sales growth rate, and foreign ownership represent firm size, financial risk, profitability, growth potential, and ownership structure, respectively. If larger size, lower financial risk, higher profitability and growth potential is expected to result in a higher company valuation, then  >0,  <0  >0,  >0 would be expected, respectively. Additionally, if foreign ownership contributes to improving corporate governance and thereby increasing company valuation, then  >0 is expected. In this study, the stickiness of R&D costs was measured using the model of Homburg & Nasev (2009). When sales in the current period decrease compared to the previous period, but the ratio of R&D expenditure to sales increases, the increase rate    is used to measure the stickiness of selling and administrative expenses.           • • where; STK: stickiness of R&D costs RND/S: R&D costs / Sales DEC: If sales decreased compared to the previous period, then 1; otherwise, 0 DEC_RND: if  〉 , then 1; otherwise0 i, t: i-th company at time t The stickiness was measured for the total R&D costs, as well as for the expensed and capitalized R&D costs separately. Each financial statement figure has been measured on a per share basis. To control for the influence of outliers on the empirical results, values of key variables such as stock price, earnings, net assets, R&D expenses, and firm characteristics that fell within the top or bottom 1% range were replaced (winsorized) with the corresponding values at the top or bottom 1%. B. Data Collection The sample for this study was selected from non-financial firms with a December fiscal year-end listed on the Korean Stock Exchange between 2012 and 2021. To ensure consistency in the information Moon Hyun Kym 83 content of financial statements, the study period was limited to after 2011, when International Financial Reporting Standards were adopted in Korea. As the stickiness of R&D expenditure is measured by the variation in the R&D expenditure ratio over time, accounting figures from 2011 were used. Non-financial firms were chosen for the study as there are significant differences in corporate characteristics between manufacturing and financial firms, and even with the same accounts, there are differences in meaning. For regression analysis using financial statement data, it is necessary to ensure the reliability of financial statement figures. Therefore, samples with inappropriate audit opinions and negative net assets were excluded. Samples were also excluded if the necessary data for variable measurement was missing or if there were less than 20 samples within the industry. The selected sample through this process consisted of 6,176 firm-year observations. Among them, 4,625 firms reported their R&D expenditures in financial statements, which accounts for approximately three-quarters (74.9%) of the total sample. Table 1 summarizes the process of sample selection. IV. Empirical Results A. Descriptive Statistics The descriptive statistics for key variables are presented in Table 2. The entire sample had an average stock price of 43,808 KRW and an average net asset Detail No. Initial observations 6,625 Less audit opinion other than an unqualified opinion, net asset<0, -113 Less missing data -21 Less number of companies within the industry<20 -315 final observations 6,176 observations with R&D costs 4,625 Table 1. Sample selection Variables Mean Std. Dev. 25% Median 75% P 43,808 84,605 2,903 11,750 41,050 E 2,265 7,070 11 444 2,103 B 40,508 78,534 3,990 10,879 38,308 AST 6.209 1.476 5.223 5.950 6.968 DR 1.027 1.362 0.296 0.669 1.226 ROA 3.290 5.973 0.790 3.120 5.970 GRW 6.212 54.928 -6.660 1.620 10.828 FR 9.812 11.855 1.550 4.770 14.818 RND 615 1,631 0 54 383 RNDE 567 1,544 0 44 341 RNDA 27 121 0 0 0 STK 0.269 1.858 0.050 0.180 0.660   0.231 1.558 0.040 0.150 0.570   0.626 1.532 0.030 0.140 0.488 Note: Sample consists of 6,176, sample with STK consists of 1,265. Variables are P: stock price, E: earnings per share, B: net asset per share, AST: ln(total asset), DR: debt/equity, ROA: operating earnings/total asset(%), GRW: sales growth ratio(%), FR: foreign equity ratio(%), RND: R&D cost per share, RNDE: expensed R&D cost per share, RNDA: capitalized R&D cost per share,   : stickiness of R&D cost,   : stickiness of expensed R&D cost,   : stickiness of capitalized R&D cost Table 2. Descriptive statistics GLOBAL BUSINESS & FINANCE REVIEW, Volume. 28 Issue. 4 (AUGUST 2023), 77-89 84 value per share of 40,508 KRW, resulting in an average price-to-book ratio (PBR) of 1.08. On average, the debt-to-equity ratio was 102.7%, the operating profit margin was 3.3%, and the revenue growth rate was 6.2%. The foreign ownership ratio was 9.8%. And, the annual R&D expenditure was 615 KRW per share, with expensed R&D costs averaging 567 KRW per share. About 92.2% of the R&D expenditure was recognized as expenses, and the amount recognized as capitalized R&D costs was not significant. The sample size for capitalized R&D costs was 1,005, accounting for 16.3% of the total sample and 21.7% of the sample that reported R&D expenses. Out of the total sample, 1,265 (20.2%) exhibited sticky behavior in R&D costs, while the sample of those who capitalized R&D costs was 268 (4.3%). When examining only the cases where R&D expenditure exhibited sticky behavior in costs, the average stickiness of capitalized R&D costs (RNDA) was 0.626, while the stickiness of expensed R&D costs (RNDE) was 0.231. This is because capitalized R&D costs are relatively more difficult to adjust according to changes in sales compared to expensed R&D costs. The breakdown of R&D expenses is reported in the footnotes, categorizing them into material costs, labor expenses, outsourcing costs, and other expenses. However, it is not easy to ascertain the specific details of the largest component, which is other expensed. Kim and Park (2019) reported that pharmaceutical and biotech companies that capitalized R&D costs did not sufficiently disclose the development stage and type of drugs in their footnotes. Table 3 presents the correlations between the key variables. The stock price demonstrated a significant positive relationship with earnings, net assets, and the firm-specific variables that were included as control variables. The debt ratio, which represents financial risk, exhibited a significant negative relationship with stock price, while firm size, profitability, growth, and foreign ownership displayed significant positive relationships. These results align with the Ohlson (1995) model, which explains firm value using earnings, net assets, and other information. The negative correlation between stock price and R&D expenditure suggests that investors may view R&D expenditure as an expense that reduces the company's value. Additionally, earnings and net assets, which are key determinants of stock prices, were found to be significantly correlated with R&D costs and most of the company-specific variables. Therefore, the stickiness of R&D costs is expected to impact the value relevance of earnings or net assets, and as these variables are also significantly related to company-specific variables, the relationship between the stickiness of R&D costs and stock prices needs to be analyzed while considering multicollinearity. B. Regression Results Table 4 presents a comparison between cases where R&D expenditure exhibited sticky behavior and cases E B RND AST DR ROA GRW FR P 0.652*** 0.775*** -0.526*** 0.465*** -0.154*** 0.251*** 0.029** 0.408*** E 0.587*** -0.285*** 0.293*** -0.217*** 0.356*** 0.039** 0.276*** B -0.384*** 0.427*** -0.161*** 0.116*** -0.103 0.288*** RND -0.371*** -0.030** -0.127*** -0.010 -0.339*** AST 0.098*** 0.213** 0.012 0.560*** DR -0.220*** -0.046 -0.130*** ROA 0.169*** 0.253*** GRW 0.004 Note: sample consists of 4,625 R&D expenditure reported in financial statements. ***, ** and * indicate statistical significance at the 1&, 5%, and 10% level, respectively. Table 3. Pearson Correlation Matrix