Today it is undoubtedly that a company should pay special attention to various types of intangibles when creating its own unique pool of resources. The reason of this lies in the following - intangibles serve as a prerequisite for a long-term competitive advantage for a common firm (Grant, 1996). What is more, researchers and practitioners emphasize importance of intangibles in terms of their influence on firm's performance (Sheehan, 2012; Archibugi et al, 2013). As a result, a growing body of academic literature is devoted to the theme of investments in intangibles.

However, it is noted that intangibles have specific features that trigger contraction of investments in these resources as a first step when a recession starts (Elexova, 2011; Sheehan, 2012). For one thing, outputs of investments in knowledge assets serve as a poor collateral for a loan, as they are mostly subject to high extent of information asymmetry (Knudsen and Lien, 2014). Furthermore, results of such investments are linked with great uncertainty (Hall, 2010). Finally, knowledge investments are widely associated with cash intensity. Necessity to apply huge amount of money in financing of knowledge investments in itself may offset all other opportunities and make a company reduce budgets related to intangibles (Cincera et al, 2012). The relative impact of these qualities of intangibles strengthens in economic turmoil as effect of demand fall and increased credit constraints of a common downturn (Knudsen and Lien, 2014).

Researchers share a common supposition that systemic accumulation of intangible assets at the time of prosperity in an economy contribute to company performance. However, as regards a period of economic turmoil, companies may balance between a bankruptcy and survival preferring to cut investment programs in a crisis (Archibugi et al, 2013; Filippetti and Archibugi, 2011). In the literature effects of different knowledge investment strategies at the time of a crisis upon future performance are not clearly explored.

The aim of the thesis is to determine whether it is more beneficial to contract or accumulate core intangibles resources at the time of economic downturn in terms of the effect on company future performance.

In order to answer the research question the database of 400 European companies is used. Methods of regression analysis helps to determine how investment decisions towards core intangibles in 2008-2009 influence companies performance in 2010-2013.

The study is organized as follows. Firstly, theoretical aspects of the theme are presented. The author observes nature of knowledge investments, as well as a body of evidence from the literature about risks and opportunities of investments in intangibles during a crisis. At the end of this part hypotheses are proposed. The next section aims to analyze companies from the sample, to be more precise - investment behavior of the companies towards intangibles. Research design and empirical results of the hypotheses testing are presented further. Finally, a reader notices concluding remarks of the work.

This thesis may be useful for those who investigate investment strategies of companies towards core intangibles and their relation to economic downturn. For those who are responsible for the process of decision making within a company this study may be interesting as well.

Now we would like to describe the sample, which we use in our analysis. The database, which serves as a basement for our sample, was collected by participants of the research group “Empirical corporate finance” (National Research University - Higher School of Economics Perm branch) during the period 2012-2014 This study comprises research findings from the «The Changing Role of Companies' Intangibles over the Crisis» carried out within The National Research University Higher School of Economics' Academic Fund Program in 2013, grant No 13-05-0021.

The parent sample represents the panel data and includes European public companies from Amadeus source of Bureau van Dijk business information service. It covers 1694 European companies from Great Britain, Germany, France, Italy and Spain for the period from 2004 to 2013. A part of this database that includes general information of the companies (country of origin, a company age and industry) as well as data from financial reports was taken from Amadeus source. Some data such as historical values of companies' market capitalization and research and development budgets was found in Bloomberg Professional service (the Terminal). Finally, some specific information that relates to intangible capital of a company was picked up in public sources of information (companies' web-sites and news agencies reports).

In order to answer the research question, the whole database of the European companies was restricted to 378 companies. The key factor for companies to fit the restricted sample was official disclosure of intangible investments (R&D and HR expenses) in annual reports in 2008 and 2009 - the period of economic turmoil. We need to mention that Italian and Spanish companies were excluded from our consideration. In the following part of this section the restricted sample of the companies will be characterized from different angles.

Geographical distribution of the companies is presented below. The majority of the sample companies are headquartered in United Kingdom and Germany (42% and 39% respectively). Slightly fewer companies have French origin (19%).

Table 3 Numerical distribution of the sample by country

It is remarkable that the age of companies from the sample averages 55 years. We might conclude that companies from the sample are mature enough. The median age of the companies is fewer - about 32. This finding does not disappear when we go deeper into the database - up to a country level: in United Kingdom, Germany and France the median age represents less values than the average age.

Table 4 Age of companies from the sample, in years

Very interesting that the oldest company from the sample is headquartered in Germany. Its name is Koenig & Bauer AG (letter C, NACE classification, manufacturing industry) and it has operated from 1817 (in 2013 its age was 196 years).

Also we need to mention that our sample does not contain newly-established firms for the period from 2004 to 2013. The youngest firm from the sample was found in 2004 (in 2013 it was 9 years old).

Table 5 contains information about the sample companies size. As to this question, we consider book value of assets and a number of employees as indicators of a company size. Referring to particular values we have evidence that average company in our sample is rather big than small. Although distribution in cases of both indicators is biased towards companies, which size is smaller than a median value, minimum value of company's assets is about €2m, and maximum is about €324bn.

Table 5 Characteristics of the size of companies from the sample, 2013

Any company from the sample employs at least 13 people. It means that the sample does not contain micro enterprises with personnel less than ten people [3, European Commission classification]. The company from the sample with the highest number of employees is Volkswagen AG (letter C, NACE classification, manufacturing industry). In 2013 it employed more than 546 thousand people.

We would like to note that values in Table 5 refer to the last year of the sample that is 2013. In Appendix A a reader may find the extended table with a country-level data about the sample companies' size.

In the analysis the authors consider the country specifics relating to innovation opportunities and efficiency of labor markets. As to this question they refer to the dataset of the Global Competitiveness report, which is published by the World Economic Forum on the yearly basis [12]. Labor market efficiency indicator from the report rises if the labor market works more effectively and is more flexible (employees hold down appropriate jobs and can shift to another economic activity rapidly and with low cost [22]). Innovation indicator regards the role of private and public sectors in research and development activities of an economy, the presence of research institutions as well as the extent of their collaboration with industries [22].

Presented diagram illustrates that the labor market in United Kingdom stayed more efficient than labor markets in Germany and France for the observed period of 2010-2013. The above diagram notes that innovation index in Germany was higher than the index in United Kingdom and France if concerning the period 2010-2013 (however the difference is not substantial).

The question of sample representativeness is crucial enough. The answer to this question reveal to what extent we might diffuse results of our analysis. In further paragraphs we will try to consider this issue by emphasizing the points of:

1. Relation between gross domestic product of the three countries and the whole product of European Union;

2. Sample companies coverage of economic activities by NACE classification;

3. A share of large firms and a group of small and medium-sized enterprises (SME) in national economies in comparison with data from population (Eurostat source of information [2]).

First of all, without a shadow of doubt the countries under consideration make a major contribution to European gross domestic product (table 6). Germany, United Kingdom and France represent the driving force of European economy comprising together more than a half of European Union GDP (about 52 per cent).

Table 6 Gross domestic product of countries under consideration, 2014 [11]

Concerning economic activities of the companies under consideration the database covers almost all economic groups presented in the European Commission classification (NACE, Rev. 2 (2008)). In Appendix B the extended table with all economic groups and corresponding distribution of the sample companies is presented.

NACE, Rev. 2 (2008) implies division of all possible economic activities by 21 groups, from A to U letters [26]. Seventeen groups out of these twenty-one have at least one firm as a representative in the sample, which constitutes about 81 percent coverage of all economic activities by the sample.

Almost one third of the sample companies (120 firms, 32%) are engaged in professional, scientific and technical activities (M group of NACE, Rev.2). Slightly fewer companies (110 firms, 29%) operate in a manufacturing sector of an economy. About one tenth of the sample (44 firms, 12%) is linked with information and communication technologies. Nine and four per cent of the sample relate to financial and insurance activities and wholesale and retail trade correspondingly (figure 7).

At the end of examination of the sample representativeness we would like to analyze companies distribution by a size class and compare it with corresponding population statistics.

For the purposes of our analysis we will divide all companies by two groups depending on a size class: (1) small and medium-sized enterprises (SME) and (2) large companies. Widespread classification implies further division of SME by micro, small and medium-sized companies [24]. But such kind of deepening will needlessly complicate examination, so that we will specify only two large groups.

By definition of European Commission SME companies should have less than 250 employees and either sales not more than €50m or book value of assets not more than €43m [24]. Large companies, consequently, should have a higher number of employed staff and bigger financial indicators. We note that in our analysis we follow denoted recommendations.

A large share of SME characterizes European economy. Relying on Eurostat information about 99.8 per cent of all companies in European Union may be regarded as SME (table 7). As for a country-level - share of SME in the whole economy is almost the same in United Kingdom and France. And only in Germany there is a slightly smaller share of SME - 99.5 per cent.

We would also like to note some remarkable values from Eurostat, which relate to statistics about population. Significant share of all European companies operate in countries presented in the sample (United Kingdom, Germany and France) - almost one third, about 29 per cent. SME from the three countries comprise the same 29 per cent in all European amount of SME. Concerning large companies - almost a half of such kind of European firms (48 per cent) are headquartered either in Germany, or in United Kingdom, or in France.

The sample distribution by size class considerably differs from corresponding population distribution (table 7). Actually about 80 per cent of all sample companies are large. For example, more than 90 per cent of German firms from the sample may be classified as big. Slightly less share of companies from France may be concerned as big firms (89 per cent). And only English firms for one third consists of SME. But it is anyway far below the level of SME share in the population (table 7).

The reason why the sample contains mainly large companies in contrary to SME is that it consists of public enterprises. It is obvious that public companies are rather big if we compare them with private ones. Therefore, we need to bear in mind that the average size of companies from our sample is big enough because of their nature (public enterprises).

In conclusion, we might say that results of our analysis may extend to large European companies (not SME). Justification for this statement may be as follows: geographically the sample covers a part of Europe where about a half of European GDP is produced and a half of European large enterprises operate.

Table 7 Consolidated statistics about SME presence in the sample and comparison with population

In the following part of the analysis we will describe dynamics of performance indicators and amount of knowledge investments for companies from the sample from 2004 to 2013.

Figure 8 demonstrates historical mean and median values of one of the performance indicators - Economic Value Added divided by book value of company's assets. We might note that 2008-2009 refers to a bottom line for this indicator for the observed period. Time before the world recession and after it: 2005-2007 and 2010-2011 periods - on the contrary are characterized with the highest levels of the relative indicator of company economic performance. The latest part of the observed period, 2012-2013, shows ambiguous dynamics of EVA/Assets, but the values anyway stay higher than ones from the bottom line of 2008-2009.

Figure 8. Dynamics of EVA/Assets of the sample companies

Another indicator of company economic performance is the return on assets (ROA, EBIT divided by book value of company's assets). Its historical dynamics is similar to one of EVA/Assets, especially when we observe the bottom line of indicators in 2009. However, ROA dynamics have some distinctive features:

(1) Median values of ROA after 2009 do not return to levels of ROA before crisis and even in 2012 and 2013 fall below the 2008 level of ROA;

(2) The mean value of ROA in 2011 goes down and reminds the dynamics of the mean value of EVA/Assets the same year. However ROA's plummeting is deeper, because in 2011 it becomes lower than it was in 2008, though 2011 is considered as a period of recovery, not a crisis.

Figure 9. Dynamics of ROA of the sample companies

Figure 10 demonstrates historical dynamics of mean and median values of R&D expenditures of companies from the sample. We need to note that the expenditures were calculated as real values of 2008 prices. The historical dynamics of R&D for the observed period looks like a hollow with a decrease from 2004 to 2005 (2006 for mean values), relatively stable values till 2009 (a bottom line for R&D) and a raise from 2009 till 2013. Particular values of R&D expenditures may be found in Appendix C, table 3.

Figure 10. Dynamics of R&D for the sample companies

Figure 11 shows historical dynamics of mean and median values of expenses on employees. All numbers are calculated as real values in 2008 prices. The mean values demonstrate positive dynamics almost in all observed years. The most remarkable growth was in 2011 and 2012. However, in 2007 the mean value of investments in human resources reported a decrease comparing with 2006 value.

Median values of expenses on employees declined in 2005 and in 2008-2009. Before and right after 2008-2009 recession the median values demonstrated steady growth.

Figure 11. Dynamics of expenses on employees for the sample companies

In the following part of the analysis we will describe growth rates of knowledge investments of the companies from the sample. What is more, we will observe different strategies of the companies related to knowledge capital contraction or accumulation in three successive periods: before, during and right after world crisis 2008-2009. investment asset strategy cost

Table 8 demonstrates annual growth rates of investments on R&D and numbers of positive and negative growth rates during the crisis. In 2008 the number of companies that accumulated R&D (increased R&D investments in 2008) equal the number of companies that contracted R&D expenses. In 2009 there was more decisions to reduce R&D budgets than to enhance (221 to reduce in comparison with 147 to increase). For information purposes we add information about growth rates for the first year of economic recovery. And numbers of contraction and accumulation of R&D budgets in 2010 was opposite to corresponding numbers of 2009: in 2010 243 companies increased R&D in comparison with 120 firms that reduce R&D the same year (see table 8).

Table 8 Descriptive statistics of annual R&D growth rates (sample companies)

Annual growth rates of expenses on employees in the period of recession produce ambiguous evidence. In the first year of recession there was more companies that contract expenses on staff. However, in the second year of crisis and the first year of recovery (2009 and 2010) companies that enhance budgets related to human resources development prevailed (see table 9).

Table 9 Descriptive statistics of annual growth rates of expenses on employees (sample companies)

The following two tables compare geometric average annual growth rates of R&D and HR budgets for three periods defined as (1) before (2004-2007), (2) during (2007-2009) and (3) right after (2009-2013) world crisis.

Average annual Growth rates and general direction of decisions about contraction or accumulation for both R&D investments and expenses on employees may be defined as similar. Average annual growth rates for both R&D and expenses on staff during the crisis are not positive, though during the periods before and after recession they are all above zero (see tables 10 and 11).

Table 10 Descriptive statistics of geometric average annual R&D growth rates (sample companies) for the three periods

General direction of decisions to enhance or reduce the investments in 2004-2013 followed some pattern: in a crisis period decisions to contract knowledge investments prevail, but during the periods of prosperity and recovery companies prefer in average increasing such kind of investments (see tables 10 and 11).

Table 11 Descriptive statistics of geometric average annual growth rates of expenses on employees (sample companies) for the three periods

The following section aims to describe how the sample companies deal with R&D and investments in human resources during the observed period 2004-2013. We again divided time horizon into three relative groups: before, during and after the world crisis. After that we examined geometric average annual growth rates of the investments in each time period. If a growth rate is positive, we assume that a company accumulate corresponding investments during the period, and vice versa. Results of such kind of analysis are summarized in the following diagrams and tables. In addition, corresponding statistics in numbers is presented in Appendix D and E.

Figure 12 shows that about two thirds of the sample companies (63%) in average increased R&D in 2004-2007. However, during the crisis the situation changed - the companies tended to contract R&D budgets and independently of the decision about R&D investments in previous time period (in two groups the share of companies that reduced R&D in 2008-2009 equals 59%).

Figure 12. Percentage distribution of sample companies among contraction or accumulation strategies relating R&D, sequence of the three periods

Table 12 Distribution of the sample companies among accumulation (^) or contraction (v) strategies relating R&D, three periods

Figure 13 shows how companies dealt in average with R&D in 2009-2013 depending on the strategy in previous years (crisis). It is interesting enough that in the group of companies that in average reduced R&D in 2007-2009 the share of companies that in average enhanced R&D in 2009-2013 is bigger than in the group of companies that in average increased R&D in 2007-2009 (76% against 62%).

Figure 13. Percentage distribution of sample companies among contraction or accumulation strategies relating R&D, sequence of two periods

Almost a half (45%) of all sample companies in average decreased R&D in crisis and improved R&D budgets during the following recovery. About a quarter of all companies (26%) in average continued to increase the investments both in crisis and in recovery period (see table 13 for details).

Table 13 Distribution of the sample companies among accumulation (^) or contraction (v) strategies relating R&D, two periods

Figure 14 demonstrates how companies dealt with investments in employees (increased or reduced) depending on the behavior in a previous period. We would like to note that the companies, which raised the budgets in 2004-2007, tended to continue increasing expenses in crisis. What is more, companies that reduced the expenses before the crisis tended to contract them in the following period.

Figure 14. Percentage distribution of sample companies among contraction or accumulation strategies relating investments in employees, sequence of the three periods

Table 14 gives us an opportunity to judge what sequence of strategies relating expenses on employees was the most widespread. And we may conclude that the bi...