A New Perspective on the Multinationality-Performance Link. Illustrate your essay with specific examples.

With their article “A Note on Return on Foreign Assets and Foreign Presence for UK Multinationals” Rugman, Yip and Jayaratne (2008) contribute to the literature, studying the relationship between multinationality and performance. Using simple statistical regression analysis, they determine positive relationship between foreign performance (measured by return on foreign assets) and the degree of foreign presence (as reported by F/T sales). This study is differentiated from previous research by using ROFA as a measure and showing its applicability to judging a company’s international competitiveness. Moreover, the article fills a gap in UK literature on multinationality and performance, which has formed for the prior fifteen years. Nevertheless, the study of Rugman, Yip and Jayaratne is somewhat limited in nature and does not provide direct insights into the usefulness of their conclusions. No connection is established with empirical evidence either and suggestions for increasing the sample size and representativeness, as well as incorporating additional factors, can be the basis for a more detailed further study on the subject.
The article examines the relationship between the degree of multinationality of firms and their performance by running a regression on data obtained from the 32 largest UK firms. Multinationality is measured by the F/T ratio (foreign/total sales), whereas performance is calculated using a variety of measures, from the more traditional ROTA (return on total assets) to the less commonly used ones ROFA and ROHA (return on foreign/home assets). This is the first research on UK international companies, which utilises ROFA to conclude that increased foreign presence has a positive effect on a company’s profitability. Using statistical techniques, the authors arrive at the following formula, expressing the linear relationship:
ROFA = a₀ + b₁ (F/T) + b₂ (SIZE) + b₃ (SIZE)² +b₄(INDUSTRY) + e_i , where INDUSTRY is a dummy variable, taking values of either 1 or 0
By using the unique approach of segmenting the performance of MNEs to home and foreign, this research allows the authors to develop statistically significant and more precise relationship, compared with prior studies.
The article includes a brief summary of the various methods and conclusions of the existing research on multinationality and performance, conducted on UK as well as on non-UK firms so far. While all prior studies are generally in agreement that multinationality improves the financial performance of a firm, they vary considerably in their methodology for reaching this conclusion. Various measures have been used to express profitability and foreign presence and while the authors do not comment on the advantages and disadvantages of each, they note that some of these prior studies have failed to establish statistically significant relationship.
Among the main strengths of the article by Rugman, Yip and Jayaratne (2008) is their attempt to revive a topic, neglected by studies of UK firms for the past 15 years, and to enrich it by offering an alternative, new method of examining the multinationality-performance relationship, via the measure of ROFA.
Certainly, the article addresses a topic that had lacked consideration with respect to UK companies for quite some time. It contributes to the conclusions of prior research by utilising more relevant, up-to-date data. The nature of businesses and their foreign operations has changed considerably over the past 15 years, due to increased globalisation and trade liberalisation. MNEs have grown in number and size and their potential for growth has been enhanced. Therefore, the authors have had a unique opportunity to test old hypotheses and their robustness in view of a new business world.
Finding an appropriate measure of firm performance and profitability seems to have been an unresolved issue among researchers. Without attempting to resolve it, the authors simply offer a new measure, which has not been used before in UK or non-UK studies. This is the return on foreign assets (ROFA), a ratio computed by dividing foreign profits by foreign assets. In contrast to a measure more commonly used before, the return in total assets (ROTA), ROFA allows for a distinction to be made between the return on assets located at home and abroad. Using a geographic-specific measure has the potential to produce more unambiguous and precise results, specifically pertaining to the performance of foreign subsidiaries. An advantage of this approach is that while ROFA is established as a primary measure of performance, ROHA and ROTA are not excluded from the analysis. The authors use all three measures, and the relationship between them, to support their study of a firm’s international performance, another feature which distinguishes their research.
A particularly beneficial side of their approach is that it allows for performance measures to be compared across companies, despite that firms vary in their definitions of accounting data. The three ratios are estimated separately, as distinct accounting practices do not allow to average for example ROFA and ROHA to obtain ROTA. As a result, a more “accurate and insightful analysis” (Rugman, 2008, p. 167) of MNEs with intensive international activities is produced.
Although the article does not directly claim that ROFA is a superior measure, Rugman, Yip and Jayaratne argue that it is particularly good in isolating the performance of foreign subsidiaries and assessing it with greater preciseness. There is no discussion, however, of the limitations of such isolation. After all, a foreign subsidiary is participating in the group’s profit and loss, and all of its activities are intrinsically related to those of the entire organisation, regardless of where it is located. Neglecting total performance at the “group level” has certain implications, the significance of which needs to be discussed. Therefore, it could be argued that one of the drawbacks that decreases the authors’ contribution to the multinationality-performance literature is its failure to find a better measure. Instead, it presents an alternative measure, which confirms results of previous studies.
Although fifteen years have elapsed since the multinationality-performance relationship of UK firms was last studied, this new research does not seem to add value to the implications of studying the topic. In other words, the claim that the authors initially make “to provide insight into the nature of competitiveness in Britain” is not in reality satisfied. Based on the regression and summary statistics, the authors directly conclude that the positive linear relationship between ROFA and F/T sales indicates that expanding of activities abroad increases profits for UK MNEs. Although it is doubtless that there is some degree of truth in this statement, there are a lot of other factors that determine a company’s success. If there are unexpected external conditions, such as the current economic and financial crisis, regardless of the degree of multinationality, a company may still be unable to make profits, in fact due to its foreign operations it may suffer even more. If most of its foreign activities are located in an area of political instability, for instance, then a crisis may have an effect on the company’s performance beyond any of the predictions of studies. So, the formula provided is neither universal, nor all-encompassing. It has its limitations and in order to discover what specific insights it provides, the analysis needs to be extended.
Among the strengths of the study undertaken by Rugman, Yip and Jayaratne is that it follows a consistent, continuous methodology, attempts to avoid the limitations of previous research and most importantly, arrives at statistically significant results.
Building a statistical model requires that great care should be taken in the collection of data: how much, how and from where it is to be obtained. The two primary measures used to assess foreign performance and foreign presence and to investigate the relationship between them are ROFA and F/T ratio. An essential part of the research is accumulating data to calculate these and it is important that complete, accurate and reliable data is obtained. The F/T ratio is measured by the exports of a company, the sales of its foreign subsidiaries and the sales at home, whereas international performance is computed using foreign sales, foreign assets and profits.
Certainly, the best type of data is the primary one, obtained through observation, experiment, questionnaires, etc. The figures used by Rugman, Yip and Jayaratne are taken from the organisational records, from the annual accounts, which in statistical analysis are considered secondary sources. Although such data is generally considered less precise and reliable than primary data, in this specific case it is the best measure that a researcher can obtain. Having in mind that on overall, financial statements are being closely audited and assumed to be free of material misstatements, it is reasonable to state that the data used to reach the conclusion of the article is reliable. Moreover, using information from the annual accounts possesses some benefits: it is already available and can be obtained relatively quickly. Although the researchers have no control over its quality, the probability for errors, invalid or inconsistent values is insignificantly small.
One of the features that distinguish the research of Rugman, Yip and Jayaratne is in fact the availability of data for assessing performance by geographic segments, i.e. ROFA. The authors mention that previous studies have been restricted in scope due to limitations caused by lack of information. Even in this study, segmented revenue data is obtainable for 33 out of the 37 UK MNEs. It does not become clear whether performing further research to include more companies would have a significant problem in obtaining this data.
In addition to the appropriateness of the data collection methods, the measures used are also found to be adequate for the purposes of the study. As already pointed out, using ROHA and ROTA in addition to the main ROFA is beneficial to the analytical part of the research. Using linear regression is also justified. The article does not elaborate much on the different stages of the statistical analysis, such as validation of the results, for example, which could be a potential weakness.
The usage of graphical representation of data is advantageous to the article as it helps visualise the concepts discussed, but some suggestions for improvement could be made. The data on foreign performance of 32 firms, presented in Table 2, could alternatively be displayed by fitting a regression line for each type of performance measure (ROHA, ROTA,ROFA) against F/T. This would be useful to the reader, demonstrating the positive linear relationship, claimed by the article conclusions. Also, the information displayed in matrix form in Figure 1 is slightly confusing, casting doubt on the significant linear fit between ROFA and F/T. Although foreign presence and foreign performance should be positively correlated, the first quadrant with high F/T and low ROFA contains the greatest number of firms. The authors do not elaborate on this. Furthermore, Figure 2 is certainly useful in illustrating the distinction between F/T, E/T and home sales but its relevancy to the statistical analysis developed is not clear. By suddenly differentiating European from Home/Foreign sales, the graph could perhaps be of value in the context of further research on the topic.
A significant part of the research on which the article is based, comes from selecting the sample size and sample content. In general, when data is analysed using statistical techniques, the appropriate size of the sample should be carefully selected, which is neither too small nor too big. Moreover, the type of companies on which regression is run should be representative, in accordance with the aims of the study. Rugman, Yip and Jayaratne are using 208 observations, which seems an adequate sample size. However, the data encompasses the sales and revenues of only 32 UK MNEs over a period of 8 years. While there are no strict rules for determining the sample size, a researcher should attempt to make reasonable judgements and be able to justify his decision. To help him in this, a wide range of techniques for selecting the appropriate sample size, depending on the specificity of the analysis, is described in Gerard van Belle’s “Statistical Rules of Thumb” (2002).
Since the authors do not discuss their methodology for data selection, it cannot be concluded whether their sample is appropriate or not. However, some suggestions that this is not a representative sample could be made. Firstly, the majority of companies contained in the sample are from the services sector, which accounts for a misrepresentation of other sectors. Secondly, firms are selected on the basis of their Fortune 500 ranking of the world’s largest firms, which means that only the biggest UK MNEs are considered. It is highly likely that the biggest companies are also the most successful ones, which implies that their multinationality-performance relationship may be stronger. The implications of this to the results of the statistical analysis are that perhaps using a more variable sample, containing MNEs with lower profitability, would have produced different and much more accurate results on all UK MNEs. Due to the limitations of the research presented in the article, the resulting regression formula is only valid for large MNEs. Rather than a misrepresentation in the sample selection, this is treated as a limitation, because the authors are clearly aware of it.
A final suggestion for improving the research in terms of sample selection relates to the usage of non-MNEs. Although the statistical analysis of foreign operations and performance cannot incorporate non-MNEs, the results of it could be compared with profitability data for non-MNEs. The authors mention that previous studies have found out that companies operating internationally outperform companies with home activities only. When they concluded that multinationality is beneficial to a firm’s performance, Rugman, Yip and Jayaratne could have strengthened their argument by comparing results with non-MNEs and showing that indeed, expanding activities to foreign markets is advantageous.
A point that needs to be considered in every research is its relation with empirical results and whether real life phenomena actually confirm the findings. The authors of the article point out that a problem with previous studies has been their failure to establish a connection with empirical evidence. No comment is made about their own research which may suggest that the problem still persists, so recommendations for including it as part of a further study seem viable.
The article “A Note on Return on Foreign Assets and Foreign Presence for UK Multinationals” by Rugman, Yip and Jayaratne (2008) makes a contribution to the literature on the multinationality-performance link, particularly for larger UK MNEs. Using ROFA, it offers an alternative approach to studying the relationship between the two measures and with effective statistical techniques, manages to demonstrate a positive correlation. Although using geographically segmented data on performance the study possesses significant potential, this potential does not seem to be properly developed. The study suffers from many limitations and restrictions, which could be overcome by initiating a more extended research on the topic. Some points to consider are increasing the sample size and variety to ensure that greater representativeness of companies, varying in size, industry and degree of multinationality is achieved. Using figures for non-MNEs could also contribute to the conclusions and applicability of the research to commenting on the actual competitiveness of UK firms. Rather than helping in the debate on how performance is to be measured, the authors extend it by offering one more option, which if further developed, could potentially produce more specific and precise results.