RELATIONSHIPS BETWEEN R&D AND PROFITABILITY: AN EXPLORATORY COMPARISON OF TWO BUSINESS SIMULATIONS WITH TWO REAL-WORLD. TECHNOLOGY INTENSIVE INDUSTRIES Developments In Business Simulation & Experiential Exercises, Volume 21, 1994 75 RELATIONSHIPS BETWEEN R&D AND PROFITABILITY: AN EXPLORATORY COMPARISON OF TWO BUSINESS SIMULATIONS WITH TWO REAL-WORLD. TECHNOLOGY INTENSIVE INDUSTRIES William C. House, University of Arkansas at Fayetteville Don M. Parks, University of Wyoming Grant L. Lindstrom, University of Wyoming ABSTRACT The purpose of this study is to provide an exploratory framework for analysis that can be used in future studies. Some of the relationships between R&D and profitability between two business simulations and two industries were used to illustrate this framework. This study explores whether R&D: profitability relationships ere the same in two industries and in two simulations. The R&D: profitability relationships for each of the four settings are examined first. The results show similarities in R&D: profitability between the two industries. For the two simulations, an equal number of similarities and differences were identified when comparing R&D and profitability. Finally, the core issue is addressed; do the BSG and MMG simulations reflect the same R&D: profitability relationships as two real world, technologically intensive industries? In most instances, the simulations do not. Questions for further exploration are then identified. INTRODUCTION One of the uses of simulations is to illustrate concepts and reinforce relationships between variables. For example, to reinforce the idea that profitability is good, business simulations reward higher profitability with higher scores. Many other relationships are embodied within simulations such as higher advertising or higher quality products able to command higher prices. Some of these relationships ere accounting-based, while others are generally accepted. Simulations must accurately reflect accounting and operating relationships, such as MARGIN x TURNOVER = RETURN ON ASSETS. Nor, would many argue that increased product quality is associated with increased price (ceteris paribus) in the real world. Simulations explicitly, and by design, reflect this type of generally accepted relationship. However, more complex, less straightforward relationships also exist in real world and simulated businesses. For example, the relationship between R&D and profitability is not uniform across industries (House & Fries, 1992). The purpose of this study is to provide an exploratory framework for analysis that can be used in future studies. Some of the relationships between R&D and profitability between two business simulations and two industries were used to illustrate this framework. It addresses the questions “What are the relationships between R&D and profitability in each of the four settings?” The next question asked is, “Are the relationships the same in each setting?” Those results are then used to compare the relationships in both simulations with the two industries. RESEARCH AND DEVELOPMENT AND PROFITABILITY Considerable attention has been devoted to measurement of the level of research and development activity and its relationship to profitability. The relationship between R&D outlays and profitability has been emphasized by Grabowski & Mueller (1988), Hirschey (1982), and Roberts & Hauptman (1987). Branch’s (1974) study of seven industries found that changes in R&D outlays ware significantly related to changes in profits. Schoeffler (1977) determined that high R&D outlays are negatively correlated with profits if the market is growing rapidly and that R&D outlays have a positive effect on performance only if the firm is in a strong position to begin with. A recent study of 727 companies for years 1983 to 1987 found that R&D intensity (i.e., R&D outlays/sales revenues) did not correlate significantly with return on sales or return on assets (Morbey & Reither, 1990). A weak relationship between research intensity and profit growth was found in computer, paper and machinery industries. In contrast, a study of growth, productivity, and profitability measures for twenty-six consumer durable manufacturing companies, twenty-six nondurable consumer products manufacturing, and twenty producer durables companies for 1991 found R&D/sales and R&D employee positively related to return on assets for the nondurable consumer companies and negatively related to return on assets for the producer durables companies (House & Fries, 1992). R&D$/Employee may be a better measure of research activity in many instances since the number of employees has lass short term variability than sales revenue. In a study of 134 companies (1978-1987), R&D/employee was found to be positively correlated with profit margin and sales per employee but not return on assets while R&D$/sales revenue was not correlated with return on sales, return on assets, or sales/employee (Morbey & Reithner, 1990). Grilches (1987) found that the level of R&D activity contributes significantly to productivity growth in larger U.S. manufacturing companies. METHODOLOGY In order to assess the extent simulations model real world effects, simulation results were compared with those for twenty-six computer hardware companies and twenty-four pharmaceutical companies for 1990. The two actual industries selected are among those considered to fall in the technology intensive category and can be expected to emphasize research and development efforts. Although it can be argued that simulation results should only be generally representative of real world outcomes, the extent to which simulation results differ from actual industry results at least gives a benchmark measure of their realism and validity. Independently, data was collected for two years for twenty-nine companies playing the Business Strategy Game (BSG) and twenty-eight companies playing a modified version of the Multinational Management Game (MMG). Both simulations are moderately complex, involving significant R&D decisions, as well as all major functional areas, including marketing, production, finance, and personnel. Quarterly data was aggregated in the MMG to permit comparisons on an annual basis with BSG. CORRELATION RESULTS-A COMPARISON OF THE IMPACT OF KEY R&D VARIABLES ON PROFIT MEASURES FOR THE TWO INDUSTRIES The first issue addressed is whether R&D: profitability relationships are the same in two industries. In a previous study of the lagged effects of productivity and R&D variables on profitability, twenty-six computer hardware/peripheral companies and twenty-f our pharmaceutical companies were selected from BUSINESS WEEK--1991 and R&D SCOREBOARD-- 199O (House & Fries, 1991). The R&D variables were correlated with return on sales and return on assets for year one and year two. As Table One shows, for the computer hardware companies, Developments In Business Simulation & Experiential Exercises, Volume 21, 1994 76 TABLE ONE COMPUTER HARDWARE COMPANIES R&D: PROFITABILITY RELATIONSHIPS ROS ROA ROS ROA YEAR 1 YEAR 1 YEAR 2 YEAR 2 R&D VARIABLES R&D $/SALES REVS-1 0.28 0.12 -0.13 0.15 R&D $/EMPLOYEE-1 0.39* 0.29 0.33 0.44 *P,0.05 N=26 R&D/employee is positively correlated with return on sales in year one and return on assets in year two. As Table Two shows, for the pharmaceutical group R&D/sales is positively correlated with return on sales end return on assets for both years but R&D/employee is positively correlated with return on sales and return on assets only in year two. TABLE TWO PHARMACEUTICAL COMPANIES R&D: PROFITABILITY RELATIONSHIPS ROS ROA ROS ROA YEAR 1 YEAR 1 YEAR 2 YEAR 2 R&D VARIABLES R&D $/SALES REVS-1 0.62 0.45* 0.71 0.56 R&D $/EMPLOYEE-1 0.35 0.17 0.68* 0.60 *P,0.05 N=24 Research intensity (i.e... R&D/sales revenues) does not seem to significantly affect profitability in the computer hardware industry, but R&D/employee affects current year return on sales and lagged year return on assets. In the pharmaceutical industry, research intensity has both a current year and lagged year impact on profitability but R&D/employee is positively related to profitability only on a lagged basis. It appears that there are significant differences in the impact of productivity end R&D variables on an industry by industry basis. Computer and Pharmaceutical Industries Compared A comparison of Tables One and Two is shown in Figure One. The computer and pharmaceutical industry samples both showed a positive correlation between all of the profitability and R&D measures except one. However, none of the statistically significant relationships are the same for both industries. This supports prior studies that suggest industry differences in the relationships between R&D and profitability (House and Fries, 1992). The next question is “what are the relationships between R&D and profitability in business simulations?” Only then will it be possible to ask whether simulations reflect the same relationships as real world industries. FIGURE ONE COMPUTER AND PHARMACEUTICAL INDUSTRIES’ R&D: PROFITABILITY RELATIONSHIPS COMPARED PHARMACEUTICAL INDUSTRY R&D/ SALES R&D/ EMP ROS-1 # # COMPUTER ROA-1 # # INDUSTRY ROS-2 ? # ROA-2 # # Both significant @ .05 with the same direction of correlation # Both same direction of correlation ? Different direction even though both may be significant @ .05 CORRELATION RESULTS—A COMPARISON OF THE IMPACT OF KEY R&D VARIABLES ON PROFIT MEASURES FOR THE TWO SIMULATIONS Table Three shows the relationship of the and R&D variables on the Income measures for the BSG companies TABLE THREE BUSINESS STRATEGY GAME COMPANIES R&D: PROFITABILITY RELATIONSHIPS ROS R0A R0S R0A YEAR YEAR YEAR YEAR R&D R&D $/SALES -0.17 -0.32 -0.54 -0.56 R&D 0.23 0.22 0.04 0.02 * P