The Market-to-Book Value Puzzle from the
Perspective of Intangible Capital: An Empirical
Comparison across Industries
Wen Chen*
Research Master Thesis 2011
Supervisor: Dr. Robert Inklaar
Co-assessed by: Professor Marcel Timmer
Faculty of Economics and Business
University of Groningen
August 2011
ABSTRACT
This paper analyzes the market-to-book value puzzle via the lens of intangible capitals (i.e. R&D stock and organization capital). Using the financial data of 192 R&D intensive firms across 15 countries and 12 industries, we find that the market-to-book ratio is reduced from 3.18 to 1.24. This improvement in solving the market-to-book value puzzle is statistically significant, meaning the puzzle is significantly less profound once intangible stocks are considered in firm’s valuation. The intangible-adjusted ratio remains somewhat larger than unity, indicating that the market participants still tend to overvalue the long-run intrinsic value of the corporations included in our sample. Moreover, these findings remain robust to a variety of alternative specifications.
Key words: Intangible capital; Firm valuation; Market-to-book value puzzle; Spillovers JEL No.: G0, M40, O30
* Research Master Student at SOM, the Graduate School of the Faculty of Economics and Business at the University
Table of Contents
I. Introduction……….1-4
II. Overview of Intangible Assets………..5-9
III. Data and Methodology……….9-17
IV. Empirical Results………..18-23
V. Conclusion and Discussion………....23-24
VI. Bibliography………...25-28
I. Introduction
With the current world gross domestic product (GDP) standing at $58259 billion (World Bank Group, 2010), the economic achievement since the Industrial Revolution has been nothing short of miraculous. Not only has the income increased exponentially, but the living standards of the people have improved significantly as well.1 An enormous
amount of research has been dedicated to studying the drives behind this economic triumph. Amongst various approaches, the sources-of-growth model has been the main empirical tool employed by the economists in tracking and explaining the growth trends. This model was initially developed in the 1950s and 1960s by Robert Solow (1956, 1957, 1960) and further improved by Kendrick (1961), Denison (1962), Jorgenson and Griliches (1967). The essence of this model is to allocate growth rates of measured output to the growth rate of labor and capital inputs, each weighted by their respective share of output, and a residual term associated with the efficiency with which output is produced from a given set of inputs. This residual is also known as Solow residual or referred to as total factor productivity (TFP). It is beyond doubt that ever since the sources-of-growth model came into existence, our understanding on economic growth has deepened sharply. Yet, despite of the popularity of this model, skepticism has been raised about our ability to understand and measure fully the fundamental sources of growth. As empirically observed, there is still a large amount of growth unexplained and a residual-based estimate is essentially a black box reflecting random shocks and various omitted variables. In Abramovitz’s words (1956), the TFP estimate is just a “measure of our ignorance”. Doubts on the sources-of-growth model were also posed by Robert Solow. Observing a slowdown in productivity that started in the late 1960s or early 1970s, Solow famously remarked in 1987: “You see computer revolution everywhere except in the productivity data”. His comment acutely pinpointed the problem inherent in the data that the economists have been using to study the sources of economic growth.
It has become increasingly clear in recent years that intangible capital, which is expensed rather than capitalized, in standard national income accounting is a key factor
1 As documented in his famous and celebrated book, Maddison (2001) noticed that over the past millennium, world
to explain the unexplained TFP residual. Parente and Prescott (2000) conclude from a variety of exercises that “unmeasured investment is large and could be as much as 50% of GDP and surely at least 30% GDP”. It is further estimated in their research that around 60% of the economy’s productivity is attributable to those unmeasured intangible capitals. Addition evidence on the importance of intangible capital is provided by Van Ark, Hao, Corrado, and Hulten (2009). They show that investment in intangibles in many advanced economies is approaching the value of investment in tangible assets, and in some cases it even exceeds tangible investment.2 With such a dramatic change in the
structure of the economy, productivity growth derived from intangible assets is likely to become increasingly important in the future (Bontempi and Mairesse, 2008). This conjecture is reaffirmed by the project carried out by OECD (2010).3 Thus, given its
increasing importance and popularity, intangible capital is at the central concern of this paper.
Although we started out with a description of the economy at the macro-level, the focus of this research and of the data are in fact on the firm-level. There are two key reasons for narrowing down the focus to the firm-level. First, the rapid expansion of investment in intangible assets is taking place at the firm-level (Corrado, Hulten, and Sichel 2006; hereafter CHS). Second, despite the fact that the majority of existing and ongoing research is centered on the aggregate importance of intangible assets, they are nevertheless, to a large extent, firm specific inputs. In other words, intangibles are (mostly) firm specific by nature and they can be most easily and directly applied to firm-level studies (Riley and Robinson, 2011).
By focusing on firms, this paper attempts to address a range of issues that puzzle the market investors about the valuations of the firms. One of the questions that have been asked repeatedly by investors is: what is the intrinsic value of a company? In a market-oriented economy, the answer to this question is simple and straightforward: a company and its assets are worth only what the highest bidder is willing to pay (Hulten
2 Countries like the United States and United Kingdom already have investment in intangibles exceeding the value of
tangible investment.
3 It is stated in its newly disclosed report that a new focus is now placed on how policies might help the accumulation
and Hao, 2008). Theoretically speaking, the holding value of the assets should be equal to the expected discounted present value of the future earnings. Like many other (economic) theories however, this one does not hold in practice. These two values can differ drastically in real life, not only during the periods of market turbulence but in calmer times as well. The value of shareholder equity is consistently valued higher by the market than the value on the balance sheet. The difference of these two values is labeled as the price-to-book gap and why such a gap exists is the market-to-book value puzzle that investors are confronted with.4
Using financial data from the companies of S&P 500 Index, the marker-to-book ratio averaged between 2.0 and 3.5 in the period of 1990-1995. During the last financial meltdown in the U.S, this ratio still remained fairly high in the range of 1.5 to 2.0 (Hulten and Hao, 2008). Therefore, the price-to-book gap is too large to be blamed solely on measurement errors of conventional equity or the vicissitudes of the stock market.
Thanks to the lead taken by-amongst others-Aboody and Lev (1998), the absence of intangible assets from corporate balance sheets is believed to be an importance source of the puzzle. Companies spend billions of dollars on research and development (R&D) as well as brand enhancement; they are however treated as current expenses by the accountants due to the lack of market transactions to measure the value of R&D and brand generated inside the company (Hulten and Hao, 2008). It is widely agreed, however, that the current success of many firms as well as their future prospects are closely associated with and even largely dependent upon their capability to develop and market the products, rather than the simple manufacturing. The example used by Michael Mander (2006) in the Businessweek Magazine could not have made this point any clearer:
Grab your iPod, flip it over, and read the script at the bottom. It says “Designed by Apple in California. Assembled in China.” Where the gizmo is made is immaterial to its popularity. It is great design, technical innovation, and savvy marketing that helped Apple Computer to sell more than 40 million iPods.
Therefore, regarding product development merely as current expenses while ignoring its contribution to future value conceals much of what makes a company successful and
4 The market-to-book value puzzle has been mirrored in the Finance literature as the disequilibrium of Tobin’s q
valuable. To this end, we aim to solve the market-to-book value puzzle through the lens of intangible assets and the main research question examined in this paper is: how much of the market-to-book value puzzle can be explained by intangibles? In addition to the efforts of unfolding this puzzle, another interesting and informative question investigated here is: do intangible assets play a significant role in determining the market value of the firm? Put differently, are intangibles really capital assets?
We extracted financial data from the Bureau van Dijk database (i.e. Orbis) and from Datastream of nearly 200 firms across 15 countries and 12 industries (see Appendix A for more information). To our knowledge, this research is the first in this field to have such an extensive coverage of countries and industries. The vast majority of existing research on this topic has placed a strong focus on firms originating from one single country, which largely restrain the generalizability of its findings.5 Given the
diversity of our dataset, the findings of this paper are believed to be more representative and reliable. In addition, the extent to which intangibles can solve the market-to-book value puzzle is also examined per industry.
We find that there is a significant reduction of the market-to-book ratio (i.e. closing of the price-to-book gap); hence the market-to-book value puzzle is significantly less profound once intangible stocks are considered. After capitalization, the ratio is just above unity (1.24), indicating that the market tends to slightly overvalue the long-run intrinsic value of the corporations included in our sample. This finding remains robust to various alternative assumptions. Moreover, with the classical knowledge spillover model, we find empirical support to the existence of positive spillovers for R&D capital.
The remainder of this paper is organized as follows; we begin in section II with an overview of the key features of intangible assets and examine how it has been applied in prior studies. In section III, we present a more in-depth discussion on the data and methods we employ to address the market-to-book value puzzle as well as the examination of the legitimacy of capitalizing intangible assets. Empirical results and main findings are presented in section IV. Section V contains conclusions and possible directions for further research in this field.
II. Overview of Intangible Assets
The concern for intangible assets and their values can be traced back many years ago, but it only gained significance and popularity very recently alongside the fundamental transformation of the structure of the economy. As shown in the graph below, our economy has transformed from agricultural to industrial and now there is a trend moving towards knowledge as a new base.
Source: Phillips, (2005)
seminar work of Corrado, Hulten an Sichel (2006) (hereafter CHS) the literature tends to agree broadly on three major sources of intangible capital.
The first group is labeled Computerized Information. According to CHS (2006), this is measured as ICT capital, consisting of software and database. The second group is named Innovative Properties. It covers more than the familiar R&D spending data as one of its major components which CHS termed it as scientific R&D. Innovative and artistic content in commercial copyrights, licenses, and designs are also included in this category which CHS termed nonscientific R&D. CHS report that, by the late 1990s, investment in nonscientific R&D was no less than the investment in scientific R&D. The last category identified by CHS (2006) is called Economic Competences. It basically represents the value of brand names and other knowledge embedded in firm-specific human and structural resources. It also gathers the expenditures designed to raise productivity and profits which are classified elsewhere other than the R&D expenses, such as brand equity which includes advertising and marketing expenditures.
Admittedly, what intangible capital exactly entails is still subject to debate and as a result of which it is hardly possible to proposal a model that describes intangibles as a particular type of good with defined economic properties. Yet, each type of intangible capital is characterized by distinctive peculiar features, there are three major attributes which are shared across different types of intangibles, namely partial excludability, interdependence and uncertainty.
intangible capital which increases the productivity of other firms that operate in similar technology domain; whereas the latter has a negative spillover effect.
The other general feature of intangibles is their interdependence (Kaplan and Norton, 2004). As described by Rastogi (2003), intangibles are “complementary, synergistic and integrative” and generate value through a complex process which involves active interaction of other intangible and tangible sources. This is also the reason why intangibles are often firm-specific or context-dependent, their value can only be fully realized for the firm that has generated them (Royal Institute of Chartered Surveyors, 2003).
Intangibles are also characterized by uncertainty, not only from the perspective of their limited temporal protection but also from the viewpoint of the process that generates them (Webster and Jansen, 2006). When firms make ex ante investment in R&D, it is not guaranteed, if at all, that this investment will bring the intended discovery. According to the research by Berndt et al. (2006), in pharmaceutical industry only 40% of drugs that start the pre-clinical process make it to the clinical stage, and only 8% of the drugs succeed to enter the market. The same uncertainty applies to the investment in job training, which aims at increasing firm’s organization capital. It is highly uncertain whether investing in training will indeed increase the level of skills of the workers and the competence of the firm as a whole.
life-cycle of the products developed by R&D. Since investing in R&D involves a large degree of uncertainty, there is also the depreciation arising from unsuccessful programs. As the other type of intangible capital that is thoroughly examined in this paper, organization capital also depreciates over time for a variety of reasons. Brand equity can lose its value if new products appear in the market place or the marketing scheme of competitors cut into its market share. In addition, organization capital can erode through worker attrition and with the adoption of new products, process and/or business models that workers are not familiar with.
It can be argued from a theoretical point of view that tangible capital depreciates quicker than intangibles because tangible assets such as machines and buildings have physical wear and tear. Put differently, physical capital is doomed to lose value through usage. On the other hand, intangible capital-such as knowledge-is immaterial and is thus not subject to depreciation through usage. Economically speaking however, capital (be it tangible or intangible) starts to lose value if there is reduction of future returns and it ceases to exist as an economic good if the returns it generates fall below the cost of producing and operating the capital (Gorzig et al., 2011). Thus, although intangible assets are not subject to physical wear and tear in the way tangibles are, they do depreciate as well due to obsolescence arising from competition. This is especially true in an information- and knowledge- explosion era in which new inventions appear on the market virtually on a daily basis, intangibles are found to depreciate much faster than tangible assets.6 In the literature, the rate of deprecation of R&D capital is estimated to
be in the range of 10% to 25% (Lev et al., 2004; CHS, 2006; Mead, 2007); for organization capital the amortization rate is found to be around 25% (Riley and Robinson, 2011; Gorzig et al., 2011). Both of which are higher than the rate assumed in depreciating physical capital, which mostly falls in the range of 1% to 15% (Fraumeni, 1997).7
6 In some sectors, such as information technology, the pace of innovation causes such rapid obsolescence that firms
have to run just to stay in place. According to estimation, computer components lose about 1 % of their value per week (Atkinson and Court, 1998).
7 According to Fraumeni (1997), some of the physical capital has a depreciate rate as high as 61% (i.e. vehicle tires).
Despite the rigorous progress and growing interest in intangible assets, not many researches are focused on addressing the market-to-book value puzzle; instead, a large amount of studies probe the relationship between intangible capital and firm performance on a micro level and economic growth on a macro level (Marrocu, et al., 2009; Hulten, 2010; CHS, 2006, 2009; Roth and Thum, 2010, Riley and Robinson, 2011). Following the lead taken by Hulten and Hao (2008) we limit our research to the investigation of the market-to-book value puzzle. Using Compustat financial data for 617 R&D intensive firms, Hulten and Hao (2008) find that conventional book value of equity explains only 31 % of the market capitalization (i.e. book-to-market ratio .31), and that boosts to 75 % if the intangible capital a firm possesses are included. Consistent result emerged from another research by Hulten, Hao and Jaeger (2009) in which they performed the same type of analysis but with R&D intensive firms coming from the US and Germany.
We follow the main guidelines suggested by Hulten and Hao (2008) in aiming to solve the market-to-book value puzzle. In the next section we proceed with the data and the methodology employed in this paper.
III. Data and Methodology
R&D spending.8 The ranking of the world’s top R&D spenders is provided by the
Department for Business, Innovation and Skills (BIS) of the United Kingdom. In total, BIS (2010) provided the ranking for 1000 firms with respect to their annual investments in R&D activities. Since these ranked firms come from different parts of the world and different industries, they are therefore classified broadly by the industries they belong to as well as their country of origin.9 Amongst those top R&D-intensive firms, BIS (2010)
identified 37 industries based on the Industry Classification Benchmark (ICB) code, of which firms selected in this paper only come from 12 (major) industries.10 For detailed
selection procedures of the industries and firms, please refer to Appendix A. Moreover, regional and industrial distribution of the selected firms can be found in table 1 and 2 (Appendix B).
Financial data compiled by Bureau van Dijk (BvD) is one of the two major sources from which the data is obtained in this paper. Data on R&D spending, book value of equity and operating surplus are retrieved from the BvD database. Since the data on stock price, selling, general and administrative expenses, firm’s market value and price-earnings ratio are either incomplete or not provided in BvD database, we relied on Datastream in retrieving those data instead. Given the fact that our data are retrieved from two separate sources, matching the data obtained from one source to the other is of key importance. For detailed combing procedures, please refer to Appendix A.
In addition, it needs to be noted that firms included in this paper come from various countries and as a result of which their accounting data are very often denominated in firms’ national currency.11 To retain comparability, all values retrieved
were converted into one single currency, namely US dollar (i.e. USD), the most widely used international currency.12
8 As a component of intangible assets, it is implicitly assumed that the more the firm spends on research and
development, the higher amount of intangibles the firm has.
9 With respect to its country origin, those firms are mostly from OECD member states: Belgium, Canada, Denmark,
France, Finland, Germany, Italy, Japan, Korea, S., Netherlands, Sweden, Switzerland, Taiwan, United Kingdom, and United States.
10 Using Industry Classification Benchmark (ICB) code, firms selected in this paper can be grouped into twelve
industries. In alphabetical order they are: Aerospace (2713), Automobile (3353), Building material (2353), Chemical (1353), Computer and software (953), Diversified industries (2727), Electronic equipment (2737), Food (3577), Industrial machinery (2757), Medical equipment (4535), Pharmaceuticals (4577), and Telecommunication (957).
11 There are only a few non-US firms reported their financial data in US Dollars.
12 Since exchange rates vary on a daily basis, we need to decide which rate to use in converting a foreign currency into
A. Can Intangibles be considered Capital Assets?
Before this paper proceeds with the capitalization of research and development outlays (R&D) and the estimation of organization capital (OC), one question that needs to be addressed a priori is: are those intangibles really capital assets? Although substantial evidence can be found in the literature supporting this view,13 with our own data set we
employ the following valuation model for validation.
Vi,t = c + β1BVi,t + β2RDi,t + β3OCi,t + β4Zi, t + εt (1)
The dependent variable Vi,t represents the market value of firm i at the end of year t,
BVi,t is the book value of the company’s equity, RD and OC are the variables of interests
in testing the legitimacy of its capitalization, and the average price-earnings ratio is denoted by Zi,t, which acts as a control variable to capture the general trend in the stock market. As usual, εt represents the stochastic error term.
With the above regression equation in mind, the current practice assumes that the coefficients β2 and β3 are not correlated with the market value, whereas the
proponents of the “new economy” or “knowledge economy” believe that these coefficients should be positive and significantly different from zero. If the latter is found to be the case (i.e. coefficients significantly different from zero), capitalizing intangibles is then legitimized in a quantitative manner. In the ensuing subsection, the legitimacy of capitalizing intangibles is also argued for from a qualitative perspective.
B. Capitalizing Research and Development Spending
After validating the treatment of intangibles as capital assets quantitatively, we are now ready to move on to the analysis of the market-to-book value puzzle. However, before this puzzle can be deciphered it is first necessary to examine whether such a puzzle does exist in our data. A quick peek at figure 1 (Appendix B) clears our concern and affirms the belief that the price-to-book gap does exist in practice, and it is observed consistently for all the years present in the data. In addition, if we take a close look at the figure, the of the firm, thus to remain consistent, for the data we obtained from Datastream we adopted the same rate that is given by Orbis.
book-to-price gap is reduced between 2008 and 2009 – the period during which the last financial crisis broke out and started to have contagious effects. It seems to be the case that the book-to-price gap is less profound during the period of crises. Nonetheless, no definitive conclusion can be drawn at this stage but it is an interesting observation which we may want to keep in mind as we proceed.
With the affirmation of the existence of the market-to-book value puzzle, we turn to address the methods that we employ to decipher it. First, we treat spending on research and development (R&D) as an investment in the company instead of a pure cost. This change of treatment of R&D expenditure is supported by Hulten and Hao (2008) with a thought experiment. As pointed out in their research, there is a fundamental asymmetry in the current accounting practice with regard to how R&D is treated. To clarify the existence of this asymmetry, consider the following scenarios.
Suppose there is a company A in the economy and it is interested in acquiring a highly technologically advanced equipment from another company. Assume that this transaction is worth $1 million. The usual economic assumption underlying this transaction is that company A invests up to the point that the purchasing cost is equal to the discounted present value of the expected future profits to be generated by this acquisition/investment. Under conventional accounting standards, the accountant of Company A will not treat this $1 million investment as an expense until figuring profit. On the other hand, the company that sells this equipment does recognize the cost of producing this investment good as an expense and adds $1 million sale to its revenues. Suppose that the actual cost of developing and producing the equipment is half-a-million; after the transaction has taken place the company that produces the equipment would then record a half-a-million before-tax profit on its income statement.
accounting practice, the $1 million would be removed from the revenue line of the income statement, because current practice treats the internal production of this innovative equipment as an intermediate input, rather than the output of an investment good. As a result, the accountant would record a current expense of half-a-million which is required for this equipment to be developed and produced; while suppressing the other half-a-million profit, that is generated by this external equipment producer like in the first scenario, into the overall operating surplus.
By considering the aforementioned two scenarios, the asymmetry in current accounting practice becomes evident: R&D is capitalized when it is produced externally but expensed when it is produced internally (Hulten & Hao, 2008). To correct for this asymmetry, R&D outlays are treated as investments in this paper and are added to the revenue line. Moreover, as can be deduced from the prior scenarios R&D outlays would also be increased by the return to the capital employed in producing the R&D (e.g. the analogue of $0.5 million profit in the example). Unfortunately, the actual magnitude of this return is unknown and we need to rely on an imputation procedure in measuring it. This procedure is developed by Hulten and Hao (2008) and allocates a total operating surplus to R&D according to the share of R&D outlays in current expenses. Therefore, the total value of Own R&D expenses can be calculated as follows:
Own R&D = R&D outlays + (R&D outlays/Current expenses) * Operating surplus (2)
After obtaining the total expenses on R&D activities there is an intermediate step needed before we can capitalize it, which is the discounting of prices. R&D investments are initially recorded in historical prices, and as prices change on an annual basis, it is necessary to discount it properly and express in constant prices in order to ensure comparability. The price deflator provided by the Bureau of Economic Analysis (BEA) R&D satellite account is used to convert nominal values into their real counterparts.
where rd t i
K, 1denotes the stock of R&D in year t-1, rd it
I represents the amount of investment in R&D in year t, and δrd is the rate of depreciation of the R&D stock.
Unlike the investment variable Ird of which the value is readily available from the financial
data we obtained, the value for the last two variables in the right hand side of equation (4) require extra attention and treatment. For the rate of depreciation, the current literature has produced a wide array of estimates ranging from 10 % to 25 % (CHS, 2006; Mead, 2007). In line with the choice of Hulten and Hao (2008), Gorzig et al. (2011) and Riley and Robinson (2011), we select an amortization procedure near the midpoint of the range suggested by the literature (i.e. δrd= 20 %). We amortize each dollar spend on R&D
over a ten-year life span, using a pattern that declines slowly in the early years and accelerates towards the end.
For the year before we first observe a firm in the data (i.e. year 2000), we follow the approach used in Gorzig et al. (2011) and Riley and Robinson (2011) to calculate the value of the starting R&D stock. In their approach the starting stocks are assumed to be proportional to the sample average of intangible investment-in this case R&D spending- in the firm. In mathematical terms it can be expressed as follows:
) 1 ( 1 ) 1 ( 1 * 2000 , g g I K rd T rd rd i rd i (4)
where T is set to 100, g is set to .02 and I-upper bar is the sample average of R&D investment. As explained in Gorzig et al. (2011), T should be infinite in theory but for practical purpose it can be set to 100. Additionally, g is the growth of investment in the years preceding the initial year and δ is simply the rate of depreciation. For detailed derivation of equation (4), please refer to Gorzig et al. (2001).
After having the R&D spending capitalized, we are now ready to adjust the conventional definition of the market-to-book ratio which is equal to the market value of the firm divided by the book value of equity. It is important to restate that the larger the deviation of this ratio is from unity (i.e. 1) the wider the firm’s price-to-book gap. The adjustment includes the own-account R&D stock in denominator. In equation:
C. Valuation of Organization Capital
As mentioned before, R&D outlays are the most widely reported form of intangible capital, but are by no mean the only type. As noticed by Danthine and Jin (2007), intangible assets may also be the result of investment in developing and launching new products, in firm’s organization capital (OC) and in human capital through (on-the-job) training and schooling. Amongst these identified types, organization capital is seen as another major component of intangible capital and its paramount importance is reflected in the extensive studies which lend support to the belief that organization capital is a significant contributor to corporate performance and growth (Arthur, 1994; Lev and Radhakrishnan, 2004; Black and Lynch, 2005; Kelly, 1996; Bailey, 1993; Dunlop and Weil, 1996). In addition, a recent research by Eisfeldt and Papanikolaou (2010) estimates that firms with more organization capital relative to their industry peers outperform those with less organization capital by 4.8 % per year. Unlike R&D outlays and other regular physical capital, however, this resource (i.e. organization capital) is rarely measured internally, nor reported publicly to investors; partially because there is no clear-cut definition what organization capital is and what does it consist of depending on the definition considered. As a result, the valuation of OC also differs14 (Black and Lynch,
2005).
There are two distinct views presented in the literature with regard to the embodiment of organization capital. Some argue that it is embodied in firm’s employees while others view it as a firm-specific capital good jointly produced with output and it is embodied in the organization itself (Atkeson and Kehoe, 2002). The proponents of the former view include Jovanovic (1979), Becker (1993), and Prescott and Visser (1980). Whereas, Arrow (1962), Rosen (1972), Tomer (1987) and Ericson and Pakes (1995) cast preference on the latter. For the purpose of this paper and in line with Lev and Radhakrishnan (2004) as well as Eisfeldt and Papanikolaou (2010), the-firm embodied - concept of organization capital is adopted. With this in mind, we follow the suggestions by Hulten and Hao (2008) and Lev and Radhakrishnan (2004) that firm’s reported selling,
14 Since the key focus of this research is not on organization capital itself, for a more comprehensive discussion on the
general and administrative (SG&A) expenses can be used as a proxy for organization capital.15 As a major income statement item, SG&A includes most of the expenditure
that generate organization capital, such as IT outlays, training cost and brand enhancement activities (Lev and Radhakrishnan, 2004). To note, although they all consider SG&A expenses as the proxy for OC, the method employed for gauging the value of organization capital differs fundamentally between CHS/HH and Lev and Radhakrishnan. The former is an imputation procedure which computes the actual value of the organization capital which firms possess; whereas the latter relies on econometric technique and provides the estimates of the annual contribution of organization capital to output growth (i.e. firm’s sales). Although it would have been preferred to have estimates for OC with a different approach, only the CHS/HH method is employed in this paper. The approach used by Hulten and Hao (2008) is similar to that developed by Corrado, Hulten and Hao (CHS, 2006) which translates their approximate proportions of brand equity and organization development investment into a corresponding fraction of SG&A spending. This approach is relatively easy to follow as it uses an imputation procedure which calculates the value of OC as 30% of the firm’s SG&A outlays. In equation, it can be expressed as follows:
Organization Investment = 30% * SG&A expenses (6)
As it is recognized by the authors themselves, estimating intangible assets using this imputation procedure is subject to some inaccuracy (Hulten and Hao, 2008). It is still employed here however for its ease of use. In addition, the presence of inaccuracy is not unique to the aforementioned imputation procedure; other proposed measurements for OC are also bound with errors and imprecision.
Similar to the need of the discounting of R&D outlays, we use the general price deflator (i.e. price deflator for GNP) provided by BEA to discount firm’s investment in
15 As shown in a study conducted by McKinsey & Company (2002) in which the financial performance of 1000 firms
organization capital and thereafter we apply the same method as in the conversion of R&D to convert investment in OC into capital stock:
oc t i oc oc it oc it I K K (1 ) , 1 (7)
ceteris paribus, from equation (4) to equation (8) only the superscripts change to OC to denote that this is capitalization for organization capital. The challenge faced here is also the valuation for the last two variables. Informed by the literature, the depreciation rate of OC is set at 25 % in this paper (Table 3, Gorzig et al., 2011). This rate is not only substantially higher than that of tangible capital but also higher than the rate used to depreciate R&D capital (i.e. 20 %). Consequently, OC capital is amortized over a shorter life time span (i.e. 6 years) while with the same pattern which declines slowly during the early phases and accelerates at the end. The logic applied in computing the starting stock of R&D is also adopted for the computation of the starting OC capital:
) 1 ( 1 ) 1 ( 1 * 2000 , g g I K oc T oc oc i oc i (8)
again ceteris paribus, going from equation (5) to (9) only superscripts change to OC while everything else remain the same.
With the computation of organization capital, the market-to-book ratio equation is further extended to the inclusion of OC capital in the denominator. In equation terms:
IV. Empirical Results
A. Regression AnalysisThe regression results of equation (1) are presented in table 4 (Appendix B) in which coefficients are reported only for those variables of interest (i.e. BV of equity, R&D stock and organization capital). As can be seen from table 4, regression is performed in both its level forms and its natural logarithms. In the latter case, the valuation model is assumed to be multiplicative.
The results shown in the first column of table 4 is obtained by imposing the restriction that intangibles do not affect market value. In this case, the estimate of equity value is2.48 and it is significant at 1% level. This implies that equity value alone is highly significant as a determinant of the market value of the firm and $1 rise in book value of equity leads to an expected $2.48 rise in market value. This result is in line with what we observe in reality–market puts a higher value on shareholder equity than what is actually recorded by the company. In column (2), two additional variables, namely R&D stock and organization capital, are introduced in the model. As can be seen, the estimate of equity value declined to 1.355, yet still significant at 1% level. This implies that after controlling for intangible stocks, equity value remains to be a highly significant determinant of market value of the firm and $1 rise in book value of equity now results in an expected $1.355 increase in market value. The same significant finding holds for the R&D stock and OC stock, both of which are found to be significant at 1%, which lend support to the capitalization hypothesis. In addition, with an estimated coefficient of 1.404, organization capital has a larger effect on firm’s market value determination than that of R&D capital (βRD=1.094). This finding conforms to the literature which labels organization capital as the most important contributor to corporate performance and growth (Lev and Radhakrishnan, 2004).
of the estimated coefficients. The coefficient show in column (3), table 4 indicates that a one percent change in equity value results in .965 % change in stock market value. This seems to suggest that there is almost a perfect one-to-one relationship between the market value and the book value and there is barely any market-to-book value puzzle left to explain. Similar conclusion can be drawn from the extended specification in which R&D stock and organization capital are added to the regression equation. As shown in column (4), the sum of their estimated coefficients is again close to unity (i.e. βSUM=1.02).
This finding is against our prior examination shown in figure 1 (Appendix C) which employs the market-to-book ratio to confirm the existence of the market-to book value puzzle.
Thus, due to the inability to address the market-book value puzzle via the estimates, we decide to follow the ‘traditional’ adjustment approach as entailed in equation (9) section III, instead. We expect to gain more insights by comparing the pre-capitalization (pre-cap) market-to-book ratio and the post-capitalization (post-cap) ratio across industries.
B. Analysis across Industries
In addition, with this intangible-adjusted ratio exceeding unity we can conclude that even if we take into account the intangible assets the firm holds, the actual value of the firm is still somewhat overvalued by market participants. In other words, the value of the companies in our sample measured according to investment cost is lower than what the market is willing to pay for the companies by 24%. According to Tobin’s q theory, this result leads us to the conclusion that on average firms included in our sample would be better off to keep on investing since the q or M/B ratio exceeds unity.
It is worth noting that the quick decline of the ratio from 3.18 to 1.24 is derived from the unweighted average of all firms. This means that if we group companies according to the industry they belong to, we may observe highly diversified patterns due to the industrial features that firms associate with. In order to gain better insights, we disaggregate the firms and re-perform the analysis industry-by-industry.
Industry Crisis of 2008-209. Moreover, with its ratio slightly smaller than unity Building material and Chemical industries also featured undervaluation of the firms.
C. Robustness Checks
In this subsection we test the sensitivity of the general results obtained earlier to different depreciation assumptions and price deflators. First, using the rate of depreciation of 20% for R&D stock and 25% for organization capital as the yardstick, we employ two alternative sets of rates to amortize R&D stock and organization capital (i.e. δRD=10% and δOC=20% VS. δRD=25% and δOC=30%). As shown in the first four columns in table 7 (Appendix B), consistent results emerge. Both R&D stock and organization capital are, again, significant determinants of firm’s market value, which lend further support to the capitalization hypothesis. In addition, as can be read from the upper and mid panels of table 8 (Appendix B), the market-to-book value puzzle is significantly less profound once intangible stocks are considered since the improvement from excluding to including intangibles is significantly different from zero. This is perfectly in line with our prior finding.
It is worth to note however, the results obtained in the upper panel, in which lower rates of depreciation were assumed, seem to be most desirable; because the post-cap market-to-book ratio is averaged .9135 which is the closest to unity compared to the other two specifications (i.e. 1.249 and 1.389) and its t-statistics has the smallest value (t-stats = 4.796). This finding leads us to conclude that intangible stocks may not depreciate as fast as we assumed, it is more accurate to amortize R&D stock over a 20-year-write-off period and organization capital over a 10-year-write-off period in solving the market-to-book value puzzle. The need to depreciate intangible capital at a slower rate is also supported by Hulten and Hao (2008). In their research, they pointed out that a 10-year-write-off period for R&D capital may be too short and it is recommended to use 20-year-write-off period as robustness check.
is again perfectly in line with our earlier findings–both R&D and organization capital are legitimized for capitalization and there is a significant closing of the price-to-book gap, hence the puzzle is significantly less profound (for statistical evidence see table 7 and 8).
D. Spillovers
As discussed in the properties of intangible capital, there will be spillovers due to its partial excludability feature. In order to examine it empirically, we employ the classical and the simplest knowledge spillover model (e.g. Jeremy Bernstein and M. Ishak Nadiri, 1989) in which the spillover pool is measured as the stock of knowledge generated by the other firms in the industry. One caveat needs to bear in mind with this model is that it assumes that firms benefit from intangible capital by other firms in their industry, and all such firms carry equal weights in the construction of the spillover. In equation terms, the extended benchmark model can be expressed as follows:
Vi,t=c+β1BVi,t +β2RDi,t +β3OCi,t +β4Zi, t+β5IndusRDi,t+β6IndusOCi,t+εt (10)
ceteris paribus, only IndusRD and IndusOC are added to the model. These two variables are computed based on the following two equations: IndusRDi j,j i ijGj and
j ij i j j i G
IndusOC , where ωij is the distance measure between firms which is assumed as one in our case for simplicity and Gj is the R&D stocks or organization
capital of other firms j.
The existence of spillovers of R&D capital can find statistical support in table 9 (Appendix B) in which the bottom row shows with all industries pooled together R&D capital feature highly significant positive spillover effect. On the contrary, no such significant effect is found for organization capital. To gain further insights, we redo the analysis by disaggregating the industries. In contrast to prior expectation, three industries feature a negative R&D spillover effect (namely Software, Electrical equipment and Telecommunication industries), of which only Telecommunication industry feature this negative spillover effect in a significant manner.
capital, with each industry examined individually more than half of the chosen industries showed negative spillover effect while the remaining featured positive effect. What is interesting is that only negative spillover effect is found to be significant. This result can be mainly attributed to the fact that the organization capital computed in this paper is a proxy based on SG&A expenses. Thus, it is no particular surprise that a significant negative spillover effect is observed for organization capital in certain industries.
It needs to be noted that two general estimates (i.e. βRD=.199 and βOC=.091)
shown in the bottom row of table 9 (Appendix B) are biased upwards, since it is implausible to assume the distance as unity for firms that come from different industries. To correct for this upward bias, a more sophisticated and rigorous distance measure is required and which is beyond the scope of this current research.
V. Conclusion and Discussion
Using financial data from nearly 200 companies across 12 industries and 15 countries, this paper attempted to decode the market-to-book value puzzle through the lens of intangible capital. We first provided econometric evidence to support/legitimize the capitalization of intangible assets. Then, we employed the conventional adjustment approach to compare the pre-capitalization and post-capitalization market-to-book ratios. As our analysis show, there is indeed a closing of the price-to-book gap (i.e. reduction in the market-to-book ratio) and hence the puzzle is significantly less profound once the intangible capital a firm possesses is taken into account. With all firms pooled together, we find a quick drop of the market-to-book ratio from 3.18 to 1.24, indicating that the market participants have the tendency to overvalue the corporations even if their intangible stocks were considered. This finding remains robust to various specifications in which alternative price deflators and depreciation assumptions for R&D and organization capital are considered. In addition, using the classical knowledge spillover model we also find there are positive R&D spillovers across firms.
firms (see table 2, Appendix B) this attempt is forfeited. Having one or two firms representing the whole country does not seem to be a plausible assumption, if at all. Thus, one direction for future research is to broaden the sample to the point that each country contains comparable amount of firms.
Moreover, although the results obtained in this paper affirm the findings of Hulten and Hao (2008, 2009), there is one major caveat need to bear in mind–there has been no adjustment for the time-value of money. As noted by DiMasi (2008), R&D investment is more complicated because of the long gestation lags. Our R&D capital estimate does not include the opportunity cost of the money tied up in the investments during this long gestation period. Admittedly, tangible assets also face the same problem, but to a much lesser degree. Problems posed by gestation lags in tangible assets are generally not of the same magnitude as the lags in intangible capital because the value of tangible investment is normally recorded at the point at which it is completed and ready to transfer to the buyer. Thus, the transaction price includes a time-value adjustment in the final cost. On the contrary, for R&D investment, what is written down on the financial statement is the cost of the R&D in progress, rather than a complete project. As a result, there is no point at which an actual market-based time-value adjustment is made. For future research it is recommended to take the time-value adjustment issue into account before capitalizing R&D investment.
This being said, the main result is unlikely to change: intangible capital is an important source for the market-to-book value puzzle and in general investors tend to overvalue the long-run intrinsic value of the firms. We end this paper with the following remark made by Al Gore (1997) in his speech at the Microsoft CEO summit:
In the Old Economy, the value of a company was mostly in its hard assets–its buildings, machines and physical equipment. In the New Economy, the value of a company derives more from its intangibles–its human capital, intellectual property, brainpower and heart. In a market economy, it’s no surprise that markets themselves have begun to recognize the potent power of intangibles. It’s one reason that net asset
BIBLIOGRAPHY
Abramovitz, Moses, 1956. Resource and Output Trends in the United States Since 1870, American Economic Review, 46 (2), 5-23
Aboody, David, and Lev, Baruch, 1998. The Value-Relevance of Intangibles: The Case of Software Capitalization, Journal of Accounting Research, 161-191
Arrow, Kenneth, J., 1962. The Economic Implication of Learning by Doing, Review of Economic Studies, 29 (3), 155-173
Arthur, Jeffery, 1994. Effects of Human Resource Systems on Manufacturing Performance and Turnover, Academy of Management Journal, 37, 670-687
Atkeson, Andrew, and Kehoe, Patrick, J., 2002. Measuring Organizational Capital, Working Paper 8722, National Bureau of Economic Research
Atkinson, Robert, D., and Court, Randolph, H., 1998. The New Economy Index: Understanding America’s Economic Transformation, Technology, Innovation, and New Economy Project, The Progressive Policy Institute
Bailey, Thomas, 1993. Organizational Innovation in the Apparel Industry, Industrial Relations, 32, 30-48
Becker, Gary, S., 1993. Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education, Chicago: University of Chicago Press
Berndt, Ernst, R., Gottschalk, Adrian, H.B., and Strobeck, Matthew, W., 2006. Opportunities for Improving the Drug Development Process: Results from a Survey of Industry and FDA, in AB Jaffe, J Lerner and S Stern eds., Innovation Policy and the Economy, Vol. 6, MIT Press for the NBER, 91-121.
Black, Sandra, E., and Lynch, Lisa, M., 2005. Measuring Organizational Capital in the New Economy, In Measuring Capital in a New Economy, Corrado C, Haltiwanger J, Sichel D (eds). Chicago: National Bureau of Economic Research and University of Chicago Press. Chicago, IL; 205-236
Bloom, Nick, and Van Reenen, John, 2006. Measuring and Explaining Management Practices Across Firms and Countries, National Bureau of Economic Research, Working paper 11948.
Bloom, Nicholas, Schankerman, Mark, and Van Reenen, John, 2007. Identifying Technology Spillovers and Product Market Rivalry, National Bureau of Economic Research, Working paper 13060.
Bontempi, Maria, E. and Mairesse, Jacques, 2008. Intangible Capital and Productivity: An Exploration on Panel of Italian Manufacturing Firms. National Bureau of Economic Research, Working paper 14108.
Brynjolfsson, Erik, and Hitt, Lorin, 2005. In Measuring Capital in New Economy, C. Corrado, J. Haltiwanger and D. Sichel, eds., Studies in Income and Wealth, Vol.65, 567-575 Chan, Louis K. C., Lakonishok, Josef, and Sougiannis, Theodore, 2001. The Stock
Market Valuation of Research and Development Expenditures, Journal of Finance 56 (6), 2431-2456
Chicago Press. Chicago, IL; 11-45
Corrado, Carol, Hulten, Charles, and Sichel, Daniel, 2006. Intangible Capital and Economic Growth. Working paper 11948, National Bureau of Economic Research Corrad, Carol, Hulten, Charles, Sichel, Daneil, 2009. Intangible Capital and US
Economic Growth. Review of Income and Wealth, (55) No.3, 661-685
Danthine, Jean-Pierre, and Jin, xiangrong, 2007. Intangible Capital, Corporate Valuation, and Asset Pricing, Economic Theory 32, 157-177
Denison, Edward, F. 1962. The Sources of Economic Growth in the United States and the Alternatives before us. New York: Committee for Economic Development
DiMasi, Joseph, A., 2008. New Drug Development in the United States From 1963 to 1999. Clinical Pharmacology and Therapeutics, 69 (5), 151-185
Dunlop, John, and Weil, David, 1996. Diffusion and Performance of Modular Production in the U.S. Apparel Industry, Industrial Relations, 35, 334-354
Eisfeldt, Andrea, L., and Papanikkolaou, Dimitris, 2010. Organization Capital and the Cross-Section of Expected Returns, Working paper, Kellogg School of Northwestern University.
Ericosn, Richard, and Pakes, Ariel, 1995. Markov-perfect Industry Dynamics: A Framework for Empirical Work, Review of Economic Studies, 61 (1), 53-82
Fraumeni, Barbara, 1997. The Measurement of Depreciation in the U.S. National Income and Product Accounts, Survey of Current Business, 7-23
Gorzig, B., Piekkola, H., and Riley, R., 2011. Production of Intangible Investment and Growth: Methodology in INNODRIVE, INNODRIVE Working paper No.1 Gu, F and Lev, Baruch, 2001. Intangible Assets: Measurement, Drivers, Usefulness. Working
paper 2003-05, School of Management Accounting, Boston University.
Hall, Richard, 1993. A Framework Linking Intangible Resources and Capabilities to Sustainable Competitive Advantage, Strategic Management Journal, 14 (8), 607-618 Hulten, Charles, R., 2010. Decoding Microsoft: Intangible Capital as a Source of
Company Growth. Working paper No.15799, National Bureau of Economic Research
Hulten, Charles, R., and Hao, Xiaohui, 2008. What is a Company Really Worth? Intangible Capital and the “Market to Book Value” Puzzle, NBER Working Paper No.14548.
Hulten, Charles, R., Hao, Xiaohui, and Jaeger, Kirsten, 2009. Intangible Capital and Valuation of Companies: A Comparison of German and U.S. Corporations, The Conference Board, New York.
Jorgenson, Dale W. and Griliches, Zvi, 1967. The Explanation of Productivity Change. Review of Economic Studies. 34, 349-383
Jovanovic, Boyan, 1979. Job Matching and Theory of Turnover, Journal of Political Economy, 87 (5), 972-990
Kaplan, Robert, S., and Norton, David, P., 2004. The Strategy Map: Guide to Aligning Intangible Assets, Harvard Business Review, Vol.32, No.5, 10-17
Kelly, Maryellen, 1994. Information Technology and Productivity: The Elusive Connection, Management Science, 40, 1406-1425
Economic Research
Lakatos, Csilla, 2010. Knowledge Capital, Product Differentiation and Market Structure: a Comparative Static CGE Analysis. Economic Seminar Paper, United States International Trade Commision Available at:
http://www.usitc.gov/research_and_analysis/economics_seminars/2010/Lakatos_Knowledge Capital.pdf
Lev, Baruch, 2001. Intangibles: management, measurement, and reporting. The Brookings Institution, Washington, DC.
Lev, Baruch, 2005. Intangible Assets: Concepts and Measurements, Encyclopedia of Social Measurement. Vol.2, 299-305
Lev, Baruch, and Radhakrishnan, Suresh, 2005. The Valuation of Organization Capital, In Measuring Capital in a New Economy, Corrado C, Haltiwanger J, Sichel D (eds). Chicago: National Bureau of Economic Research and University of Chicago Press. Chicago, IL; 73-99
Maddison, Angus, 2001. The World Economy: A Millennial Perspective. Paris: Development Center of OECD.
Mandel, Michael, 2006. Why the Economy Is A Lot Stronger Than You Think. Businessweek Magazine, February 13, 2006. Available at:
http://innovate.typepad.com/innovation/files/measuring_the_innovation_econo my_bw_article_feb_13_2006.pdf
Marrocu, Emanuela, Paci, Raffaele, and Pontis, Marco, 2009. Intangible Capital and Firm Productivity. Working paper CRENoS 2009-16. University of Cagliari.
Mead, Charles, I., 2007. R&D Depreciation Rates in the 2007 R&D Satellite Account, Bureau of Economic Analysis/National Science Foundation, R&D Satellite Account Background Paper, U.S. Department Of Commerce
Organization for Economic Cooperation and Development (OECD), 2010. New sources of growth: intangible assets. Available at
http://www.oecd.org/dataoecd/60/40/46349020.pdf
Parente, Stephen, L. and Prescott, Edward, C., 2000. Barriers to Riches. New York: The MIT Press.
Phillips, Jack, J. ,2005. Investing in your Company’s Human Capital: Strategies to avoid spending too much or too little. New York: American Management Association
Prescott, Edwards, C., and Visscher, Michael, 1980. Organization Capital, Journal of Political Economy, 88 (3), 446-461
Rastogi, P.N., 2003. The Nature and Role of IC. Rethinking the Process of Value Creation and Sustained Enterprise Growth, Journal of Intellectual Capital, Vol.4, No.2, 227-248
Riley, Rebecca, and Robinson, Catherine, 2011. UK Economic Performance: How Far Do Intangibles Count? INNODRIVE Working paper No.14
Roth, Felix, and Thum, Anna-Elisabeth, 2010. Does Intangible Capital Affect Economic Growth. Working document No.335, Center for European Policy Studies.
Royal Institute of Chartered Surveyors, 2003. Valuing Intangible Assets: a report by the Center for Business Research, London
Journal of Economics, 70, 65-94
Solow, Robert, M. 1957. Technical Change and the Aggregate Production Function. Review of Economics and Statistics, 39, 312-320
Solow, Robert, M. 1960. Investment and Technical Progress. In Mathematical methods in the social sciences, ed. K. Arrow, S. Karlin, and P. Suppes, 89-104. Stanford, CA: Stanford University Press
Solow, Robert, M. 1987. Review of The Nature of Economics, by Jane Jacobs. New York Times, July 12, 36
Stewart, Thomas, A., 1997. Intellectual Capital: The New Wealth of Organizations, New York: Bantam Doubleday Dell Publishing Company
Tomer, John, F., 1987. Organizational Capital: The Path to Higher Productivity and Well-being. New York: Praeger Publishing Company.
Uppenberg, Kristian, 2009. Innovation and Economic Growth. European Investment Bank, Working paper series Vol.14, 10-32
Van Ark, Bart, Hao, Janet, Corrado, Carol, and Hulten, Charles, 2009. Measuring Intangible Capital and its Contribution to Economic Growth in Europe. European Investment Bank, Working paper series Vol.14, 62-88
APPENDIX A
Data Description
Selection of the industries
The remaining industries are excluded from the analysis simply because the number of firms classified in those industries is too small to be included in the analysis. For instance, there are only two firms in the Delivery Service industry (ICB: 277) and three firms in the Renewable Energy Equipment industry (ICB: 583). In comparison, 112 firms can be found in the pharmaceutical industry. In order to have a comparable sample size of the firms across industries, we decide to discard those industries that have an insufficient amount of firms. Given 1000 firms and 37 industries identified accordingly, each industry should, on average, contain 27 firms (i.e. 1000/37). In order to have a broader coverage of the industries, we decide to put a 20% lower bound on the average value computed earlier as the cut-off point (i.e. 0.8*27). In other words, only industries that contain strictly more than 22 firms are considered. This attempt saved Building Materials industry (ICB: 235), which has 23 firms in its group, from exclusion.
Selection of the firms from the chosen industry
narrowing down the sample in such a manner, the firms we choose to include is now purely dictated by data availability. In the end, the sample size is restricted to 192 firms and it is collected for the time period from 2001 to 2010 (i.e. a 10-year time span).
Combining the data sources
APPENDIX B
Table 1 – Industry List
Table 2 – Country List
Table 3 – Breakdown of the Broadly Identified Industries Industry 1: 2713 Aerospace 2717 Defense Industry 3: 3353 Automobiles 3355 Auto parts 3357 Tires
Industry 6: 1353 Commodity Chemicals
1357 Special Chemicals
Industry 7: 2353 Building Material
2357 Construction
Industry 9: 2737 Electronic Equipment
2733 Electric Components
Industry 14: 3577 Food Products
3573 Farming and Fishing
Industry 17: 2727 Diversified Industries
Industry 19: 4535 Medical Equipment
4537 Medical Supplies
Industry 21: 2757 Industrial Machinery
2753 Commercial Vehicles and Trucks
Industry 34: 4577 Pharmaceuticals 4573 Bio-technology Industry 37: 9537 Software 9533 Computer services Industry 39: 9578 Telecommunications&Semiconductors 9572 Computer Hardware
ICB code Description
Table 4 – Determinants of Firm’s Market Value LEVELS LOGS (1) (2) (3) (4) δRD=.2 & δOC=.25 δRD=.2 & δOC=.25 Equity 2.482*** (19.18) 1.355*** (8.07) .965*** (65.56) .6771*** (22) R&D 1.094*** (3.68) .1438*** (5.28) Org. Capital 1.404*** (4.28) .2031*** (6.08) Constant 94916 (.04) -7430514 (3.6) 1.473*** (5.83) .5** (2.49) Obs 1915 1915 1915 1915 R2 .723 .79 .879 .897
Note: the regression results also include country and industry fixed effects, which are intentionally omitted in this table. P-E ratio is also omitted since it acts as a control variable and is not of interest to interpret. Absolute values of t-statistics based on robust standard errors are reported in parentheses. Moreover, the results in columns (1) and (2) refer to specification in which the variables are in levels. The results in columns (3) and (4) refer to specification in which market value, R&D stock, and organization capital is transformed to natural logarithms. *** p<0.01, ** p<0.05, * p<0.1
Table 5 – Statistical Testing of the Improvement of the Market-to-Book Puzzle
aHo:mean=1
bHo: mean=0 Obs 1915
Mean Std.Err t-Statistics
δRD =.2 aPre-capitalization 3.189 0.0992 22.063***
& aPost-capitalization 1.249 0.0229 10.876***
δOC =.25 bPostcap – Precap -1.94 0.0897 -21.612***
Table 6 – Analysis per Industry
Precap ratio Postcap ratio
Aerospace*** Mean 3.277 1.285
t-stats 7.088 4.441
Automobile*** Mean 5.8 0.568
t-stats 5.801 -20.784
Building Material** Mean 1.757 0.867
t-stats 8.841 -2.69 Chemmical*** Mean 1.792 0.915 t-stats 17.287 -3.377 Software*** Mean 4.454 1.726 t-stats 17.804 9.122 Diversified*** Mean 2.865 1.221 t-stats 10.227 3.161
Electric Equipment Mean 2.097 1.019
t-stats 14.405 0.418
Food*** Mean 4.751 2.589
t-stats 5.899 3.972
Industrial Machinery Mean 3.081 1.081
t-stats 3.407 1.407
Medical Equipment*** Mean 6.684 2.179
t-stats 7.216 12.354
Pharmaceutical*** Mean 4.119 1.33
t-stats 12.92 9.146
Telecommunication** Mean 3.084 1.164
t-stats 12.483 2.315
Note: t-stats is computed to test whether the market-to-book ratio of the industry under concern is significantly different from one.
Table 7 – Robustness to Determinants of Firm’s Market Value
LEVEL LOGS LEVEL LOGS LEVEL LOGS
(1) (2) (3) (4) (5) (6)
δRD =.1 & δOC =.2 δRD =.25 & δOC =.3 Alternative price deflators
Equity 1.33*** (8.67) .689*** (22.72) 1.372 *** (7.94) .68*** (22.02) 1.348*** (8.3) .686*** (22.21) R&D .686*** (5.03) .151*** (6.02) 1.223*** (3.18) .151*** (5.49) 1.23*** (3.71) .161*** (5.69) Org. Capital .838*** (4.49) .184*** (5.76) 1.794*** (4.32) .189*** (5.77) 1.28*** (3.62) .173*** (5.3) Constant -8717914 (4.73) .288 (1.42) -7039943 *** (3.36) .603*** (3) -8377849 (4.05) .503 (2.5) Obs 1915 1915 1915 1915 1915 1915 R2 .798 .897 .787 .897 .795 .896
Note: the regression results also include country and industry fixed effects. Absolute values of t-statistics based on robust standard errors are reported in parentheses. Moreover, the results in columns (1) and (2) refer to specification in which the depreciation rate of R&D capital is assumed at 10% and that of organization capital is at 20%. The results in columns (3) and (4) refer to specification in which R&D stock depreciates at 25% and organization capital at 30%. Columns (5) and (6) are the results obtained with an alternative price deflator provided by Bureau of Economic Analysis (i.e. implicit price deflator for GDP). *** p<0.01, ** p<0.05, * p<0.1
Table 8 – Robustness to Statistical Testing of the Improvement of the Market-to-Book Puzzle
aHo:mean=1
bHo: mean=0 Obs 1915
Mean Std.Err t-Statistics
δRD =.1 aPre-capitalization 3.189 0.0992
0.0175 0.0929
22.063***
& aPost-capitalization .9135 4.796***
δOC =.2 bPostcap – Precap -2.273 -24.469***
δRD =.25 aPre-capitalization 3.189 0.0992
.02512 .00832
22.063***
& aPost-capitalization 1.389 15.495***
δOC =.3 bPostcap – Precap -1.8 -20.38***
Alternative aPre-capitalization 3.189 0.0992
.02298 .0897
22.063***
Price deflators aPost-capitalization 1.248 10.8***
bPostcap – Precap -1.932 -21.547***
Table 9 – Spillovers within the Industry βRD Std.Err. t βOC Std.Err. t Aerospace .168 .0696 2.42** -.5022 .5044 1 Auto 1.01 .7234 1.4 -2.57 1.492 1.73* Bld.Mat .2859 .0761 3.75*** -1.946 .6175 3.15*** Chemicals .5456 .8534 .64 1.659 .902 1.84** Software -.7479 .7009 1.07 .0214 .8065 .03 Diversified 1.197 1.772 .068 1.468 1.773 .083 Elect.Equip -1.203 .7643 1.57 -1.031 .8628 1.2 Food .2693 .8458 .32 -1.482 .6274 2.36** Indus.Machi 1.525 1.032 1.48 -4.396 1.782 2.47** Med.Equip 3.051 1.243 2.45** -3.063 1.225 -2.5** Pharm .2051 .7265 .28 .934 1.029 .91 Telecomm -2.737 .6183 4.43*** .8138 .6978 1.17 Average .199 .037 5.38*** .091 .1041 .88
Note: R&D stocks and organization capital are transformed to natural logarithms for regression. Absolute values of t-statistics are calculated based on robust standard errors.
