Economic return on investments in research and development (R&D) in Pinus and Eucalyptus plantations in Brazil

Currently, public and private resources are directed towards the development of Research and Development (R&D) projects in the Brazilian forestry area. But, many times, such investments are used only as corporate advertising, without the knowledge of their real return, underestimating the importance of R&D in the development of organizations in this sector. Hence the importance of studies that seek to evaluate investment returns and how profitable they are for society and for technological innovation. The objective of this study was to evaluate the possible economic return of R&D in the Brazilian Forest Sector, in particular, its effects on increasing the productivity of Pinus and Eucalyptus stands. Through the publications of the Statistical Yearbook of the Brazilian Association of Planted Forest Producers (ABRAF) it was possible to understand the relationship between investments in R&D, planted area and annual current wood productivity, using an indicator that related Investments and Revenue in the R&D of organizations Brazilian forestry companies of this Association. It can be inferred that for every R$ 1.00 invested in R&D projects, an average return of R$ 15.02 is obtained in such organizations. It was also concluded that the marginal gains in annual wood productivity correlated positively and significantly with investments in R&D (r = 0.43). Thus, investments in R&D made by silvicultural organizations translate into an economic return for them, as well as that their absence can stagnate the increase in wood productivity. Retorno econômico dos investimentos em pesquisa e desenvolvimento (P&D) em plantações de Pinus e Eucalyptus no Brasil RESUMO: Atualmente, são direcionados recursos públicos e privados para desenvolvimento de projetos de Pesquisa e Desenvolvimento (P&D) na área de silvicultura brasileira. Mas, muitas vezes, tais investimentos são utilizados apenas como propaganda corporativa, sem o conhecimento do seu real retorno, subvalorizando-se a importância da P&D no desenvolvimento das organizações deste setor. Daí a importância de estudos que procurem avaliar os retornos dos investimentos e o quanto são proveitosos para a sociedade e para inovação tecnológica. O objetivo deste estudo foi avaliar o possível retorno econômico da P&D no Setor Florestal Brasileiro, em especial, seus efeitos no aumento da produtividade dos povoamentos de Pinus e eucalipto. Através das publicações do Anuário Estatístico da Associação Brasileira de Produtores de Florestas Plantadas (ABRAF) foi possível entender a relação entre investimentos em P&D, área plantada e produtividade de madeira corrente anual, usando-se um indicador que relacionou Investimentos e Receitas na P&D das organizações florestais brasileiras desta Associação. Pode-se inferir que a cada R$1,00 investido em projetos de P&D, obtém-se um retorno médio de R$15,02 em tais organizações. Também se concluiu que os ganhos marginais de produtividade (anuais) de madeira correlacionaram-se positivamente e significativamente com os investimentos em P&D (r = 0,43). Assim, os investimentos em P&D realizados pelas organizações silviculturais se traduzem em retorno econômico para as mesmas, bem como que a ausência deles pode estagnar o aumento da produtividade de madeira. Original Article *Corresponding author: viniciusesilva@yahoo.com.b

ABSTRACT: Currently, public and private resources are directed towards the development of Research and Development (R&D) projects in the Brazilian forestry area. But, many times, such investments are used only as corporate advertising, without the knowledge of their real return, underestimating the importance of R&D in the development of organizations in this sector. Hence the importance of studies that seek to evaluate investment returns and how profitable they are for society and for technological innovation. The objective of this study was to evaluate the possible economic return of R&D in the Brazilian Forest Sector, in particular, its effects on increasing the productivity of Pinus and Eucalyptus stands. Through the publications of the Statistical Yearbook of the Brazilian Association of Planted Forest Producers (ABRAF) it was possible to understand the relationship between investments in R&D, planted area and annual current wood productivity, using an indicator that related Investments and Revenue in the R&D of organizations Brazilian forestry companies of this Association. It can be inferred that for every R$ 1.00 invested in R&D projects, an average return of R$ 15.02 is obtained in such organizations. It was also concluded that the marginal gains in annual wood productivity correlated positively and significantly with investments in R&D (r = 0.43). Thus, investments in R&D made by silvicultural organizations translate into an economic return for them, as well as that their absence can stagnate the increase in wood productivity.

Introduction
Lee (2020) begins his works by warning that productivity is not equivalent to product or production. According to this author, productivity reflects improvements in the ability to transform inputs into products. In the most literal sense, it is a residual measure of the contribution to product growth after all factors have been considered. It is the non-physical product of innovation, efficiency, management, research, climate and luck. In this way, the economic development of companies is directly related to their intellectual capital, with intellectual capital being the name given to all information, transformed into knowledge that adds to those that you already have.
According to Wen (1993) and Garcia-Marin and Voigtländer (2019), there are three factors that drive productivity: increase in the quantity of inputs, institutional innovation, and the third source of growth is technological progress, which shifts the production function upwards. Technological progress is generally embedded in improving human quality and physical capital. Therefore, with the same number of inputs, more product becomes available.
Many public and private resources are directed towards the development of Research and Development (R&D) activities, but they are often used only as corporate advertising, and, thus, the importance of R&D in the development of Brazilian Forestry companies is underestimated.
Especially at the time of the great financial and economic crises that afflict companies, the resources directed to R&D activities are reduced, as they are treated as investments in the financial structure of companies, and not as direct inputs of production. With that, several research are interrupted and the field of knowledge goes through a regressive state.
Consequently, it is essential to carry out studies that seek to evaluate the returns on investments in research and to determine how profitable these have been proving to society. A limited number of studies are found in the literature aiming to measure the economic returns from investments in Research and Development -R&D in different productive sectors. Especially, in the case of agriculture, there are many studies of this nature, such as Huffman and Evenson (1993), Alston et al. (1998), Ruttan (1982), Fuglie et al. (1996) and Yee et al. (2002), who evaluated the returns to R&D in North American agriculture; Rosegrant and Evenson (1992), who assessed the returns to R&D in agriculture in South Asia; Gasques et al. (2000), who evaluated the returns of R&D in Brazilian agriculture; Araújo et al. (2002), who evaluated the returns to R&D in agriculture in São Paulo, Figueiredo et al. (2012) reduced the R&D gains in the São Paulo citrus industry in observed, significantly increasing the leverage of R&D in citrus culture.
Brazil has a wide vocation in the forestry area and is currently one of the most competitive countries in terms of productivity and cost of wood, largely due to the high level of forestry and industrial technology, provided by R&D activities (ABRAF, 2013).
In this context, the objective of this work was to evaluate the economic return of investments in Research and Development (R&D) in the Brazilian Forest Sector, in particular, it's effects in increasing the productivity of Pinus and Eucalyptus forests, and indirectly in reducing wood costs.
The quantitative analysis presented in the Yearbooks derived from the collection of primary and secondary data. Primary data were obtained through: • Completion of complete questionnaires by ABRAF member companies; • Completion of simplified questionnaires by state collective associations associated with ABRAF; • Contact with class associations such as ABIPA ( Information on planted areas in Brazil was presented by state of the Federation, and the methodology for obtaining these estimates was detailed below, for each state: • Amapá: the planted area was estimated from direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 22.5% more or less. • Goiás: the planted area was estimated from direct contact with forestry companies not associated with ABRAF and through the mapping of forestry plantations using Landsat -5 satellite images (2011/2012), with a spatial resolution of 30 meters. It is estimated that the margin of error can vary by 20% more or less.
• Bahia: the planted area was estimated based on information made available by individual members to ABRAF and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error may vary by 7.4% more or less.
• Espírito Santo: the planted area was estimated based on the information provided by ABRAF's individual associates and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error may vary by 14.3% more or less.
• Maranhão: the planted area was estimated based on the information made available by ABRAF individual members and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 12.2% more or less.
• Pará: the planted area was estimated based on the information provided by ABRAF's individual associates and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 26.6% more or less.
• Mato Grosso: the planted area was estimated based on information made available by individual and collective associates to ABRAF, as well as by direct contact with non-associated forestry companies. It is estimated that the margin of error can vary by 16.4% more or less.
• Mato Grosso do Sul: the planted area was estimated based on information provided by ABRAF individual members, REFLORE and direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 3.3% more or less.
• Minas Gerais: the Eucalyptus planted area was estimated based on the information provided by AMS in view of the information obtained by individual ABRAF members and by direct contact with companies not associated with ABRAF. The planted area of Pinus was estimated based on the information made available by individual associates to ABRAF and by direct contact with non-associated forestry companies. It is estimated that the margin of error may vary by 2.1% more or less.
• Paraná: the planted area was estimated based on information made available by individual members to ABRAF, by the Paranaense Association of Forest Based Companies (APRE), by the Association of Leaf Smoke Planters in Rio Grande do Sul (AFUBRA), and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error may vary by 9.8% more or less.
• Rio Grande do Sul: the planted area was estimated based on information provided by individual members to ABRAF and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 16.4% more or less.
• Santa Catarina: the planted area was estimated from information provided by individual ABRAF members, by collective members of the Catarinense Association of Forestry Companies (ACR), APRE and AFUBRA, as well as by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 16.4% more or less.
• São Paulo: the planted area was estimated based on the information provided by ABRAF's individual members and by direct contact with non-ABRAF forestry companies. It is estimated that the margin of error may vary by 10.1% more or less.
• Tocantins: the planted area was estimated based on information made available by individual members to ABRAF, by the Association of Tocantins Reforesters (ARETINS), and by direct contact with non-ABRAF forestry companies. It is estimated that the margin of error may vary by 9.8% more or less.
• Piaui: the planted area was estimated based on the information made available by ABRAF individual members and by direct contact with forestry companies not associated with ABRAF. It is estimated that the margin of error can vary by 8.2% more or less.
• Other states: the area planted in the other states was estimated through the compilation of information provided by individual members to ABRAF and direct contact with forestry companies not associated with ABRAF.
The error statistics were obtained through data variability, the greater the variability, the greater the statistical error (ABRAF, 2006(ABRAF, to 2013 (%) . = accumulated inflation in the period 2006 to 2020 given in percentage.
= annual inflation observed as a percentage in Brazil in the umpteenth year.

Calculations to Assess R&D Gains
To assess the average productivity of wood in Brazil, the average for each Brazilian state of Eucalyptus and Pinus species was considered, weighted according to the planted plant areas, according to the weighting methodology proposed by Filho et al., (2019), and described in the equation below: On what: . = average productivity weighted by planted area (ha). = productivity of Pinus or Eucalyptus in the umpteenth state of Brazil.
= planted area of Pinus and Eucalyptus in the umpteenth Brazilian state.
The average price of Pinus and Eucalyptus wood was also calculated weighted by the area planted by each species in the territory of Brazil, according to the equation below: On what: . = average price of wood weighted by planted area (ha).
= price of wood of Pinus or Eucalyptus in the umpteenth state of Brazil.
= planted area of Pinus and Eucalyptus in the umpteenth Brazilian state.
In order to obtain a clear cause-effect relationship between investment and revenue, the investments of the companies that were basically the companies associated with ABRAF and those described above in this article were counted, with the forest product considered by the companies basically wood, Pinus or Eucalyptus. In other words, of the 27 companies associated with ABRAF, only 1 has Tectona grandis wood as its main, which is responsible for only a small part of the investment between ABRAF member companies. However, the investments made by Floresteca were removed from the analysis to avoid confusion in the analysis of the data, which would not make sense. In this sense, the calculation of investments made by companies was carried out according to the equation below: . ( $) = ∑ After having all the data of average wood productivity organized on an annual basis, it was possible to calculate the gains obtained from year to another year (marginal gain), relative to the hectare, given according to the equation below: Consequently, the financial gains related to wood productivity gains in reais were calculated through the product between the annual marginal productivity gain and the market price of Pinus and Eucalyptus wood, according to the equation below: ( With this information, it was possible to establish an index of investment in R&D and economic return (I / R), which in accounting is commonly called Return on Investment (ROI). The simplest expression of ROI, which is applied to information extracted from accounting, is given by the following equation: On what: / = investment on return index. = investment made by the nth forestry company in Brazil. $ = financial gain obtained by investments in research from 2006 to 2013 in Brazil. Pearson's product moment correlation coefficient was used for Box-Cox transformed data with statistical significance determined with the Student's t-test (p < 0.05). Linear correlations were performed for the variables: average area planted (ha), . ..The angular coefficients of the linear equations generated by Pearson's correlations were analyzed to statistically visualize the tendencies of increasing or decreasing each of these variables (THIBOS et al., 2002).

Results and discussion
It can be seen in Figure 1 clearly that the area planted with Eucalyptus (AE) undergoes an increase in the years evaluated, and the area planted with Pinus (AP) goes through the reverse process. This can be verified by the angular coefficient of the AP line that is positive (angular coefficient = 254,539 x), that is, the trend of increasing the planted area over the years in Brazil, and with AP, the trend is negative (-32,706 x). One of the explanations of the managers of the political bodies linked to the forestry area for this fact is that it is due to the change in the direction of the products to the pulp market with only Eucalyptus wood, and the unsuitability of Pinus wood for pulp production; as well as the production of energy through charcoal (ABRAF, 2014), which the Pinus wood is not suitable for this purpose.
Above all, it is clear that the total forest area planted in Brazil with vaious species increases every year, with the effective contribution of forests planted with Eucalyptus. The proportional contribution of timber supplied from natural forests has steadily declined to less than half of global demand in 2015, with current harvest rates for natural forests declining since 1995. This shift in dependence on forest plants to supply demand for forest products wood resulted from the increase in the area dedicated to planted forests (50% increase between 1990 and 2015; Payn et al., 2015) combined with accelerated growth rates. Planted forests represent about 7% of all forests (Whiteman, 2014). The fastest growing and most intensively managed plantations are responsible for only 1.5% (54 million ha) of the world's forests, but provide a third of the supply of non-firewood (INDUFOR, 2012). On the other hand, the gains in productivity have been smaller since 2009, with the absolute gains not exceeding 0.5 m³ ha-1 year-1 after 2009 (2012 to 2013), whereas before from 2009, gains of up to 3.4 m³ ha-1 year-1 were obtained, which was the case for productivity from 2006 to 2007 ( Figure  2). As the Eucalyptus crop has a long cycle compared to the cycle of other annual crops, it cannot be to state that the losses / gains in wood productivity in their respective harvest years occurred as a result of silvicultural management Linear (Total area planted with pine (ha)) Linear (Total area planted with eucalyptus (ha)) practices adopted in the same year in which the wood was harvested. According to Hakamada et al. (2015) silvicultural management has the function of increasing the availability of growth resources, such as water, light and nutrients (Nyland, 2007). Studies carried out with forest plantations that explored improvements in soil preparation practices Allen, 2003), fertilization (Nilsson et al., 2002), spacing adjustment (Silva, 2006) and control of weeds (Little et al., 2003), showed an increase in the uniformity and productivity of wood when there was an increase in the availability of resources, as a result of less competition between individuals (Binkley et al., 2002), however this occurs throughout of a cycle of 5, 6, 7 years, or more, in the case of Pinus, which is the equivalent of a rotation in Brazil today. This was a methodological limitation of this work, since it is not identified exactly when the gain was realized, that is, it is not known throughout the cycle when the Technology was used, however it is intended here to demonstrate general trends of increase / decrease investments in R&D and their respective impacts on productivity, and not exactly measuring the annual gains in these indexes. Over the period 2006-2020, Brazilian inflation behaved in a heterogeneous manner in terms of the distribution of values over the period studied ( Figure 3). This period was marked by high growth in international commodity prices, nominal exchange rate volatility and different phases of the influence of unit labor costs on the Brazilian inflationary process (Gomes and Freitas, 2020).   Figure 4 shows the rapid devaluation of the average price of wood in Brazil in the period evaluated, a fact seen when analyzing the data angular coefficient of the line generated with the data (R² = 0.8318), and angular coefficient of -2.3058). According to Almeida et al. (2019) wood prices in Brazil, despite showing high and low movements over the years, have had constant decreases. The author also concludes that analyzing the historical series of the real price of Eucalyptus, even without inflation, we noticed that the value of the product decreased, while the prices of Pinus were more stable over the period. The volatility in wood prices, with frequent price reductions, were the result of high wood stocks and crises in the specific charcoal sector, which is totally linked to a crisis in the steel sector (Miranda et al., 2014).
Another point is the fact that inflation in the period 2006 to 2013 rises rapidly in Brazil, especially after the year 2008 ( Figure 4). This shows a negative and unfavorable financial and economic scenario for wood producers, as it shows a scenario in that: i) company revenues fall due to the fall in wood prices, ii) production costs go up as demonstrated by the rise in inflation over the years, and automatically, iii) the profit margin is pressed down and the forestry business it becomes unattractive. Another fact that further aggravates the delicate situation of Brazil's wood producers is the issue that the inflation of forest production measured by the National Index of Costs of Forestry Activity (INCAF-Poyry) is generally above the average inflation in Brazil. For example, in 2012, inflation in the Brazilian forestry sector was 14.7% per year (p.y.), three times higher than the IPCA (5.8% p.y.) and four times higher than the average international inflation (4.0% p.y.) (ABRAF, 2013). This overheated inflation in the forestry sector be due to several factors, such as, for example, the increase in the prices of diesel oil and imported fertilizers, due to the appreciation of the international price of the barrel of oil and the dollar over the years of the evaluated period (Chernov and Sornette, 2020). In this sense, the increase in the exchange rate influences the increase not only of fertilizers and diesel oil, but also of several inputs used in the forest production area that are imported, such as insecticides, fungicides, herbicides, which are imported from several countries of the world. Investments in R&D made by Brazilian forestry sector decreased during the period evaluated, especially in 2009, as shown in Figure 5. This reduction in investments in R&D was due to the 2008 crisis, caused by the high real estate speculation in the United States of America (USA). According to Guttmann (2015) this crisis was chained by the high investment of the US government in war policies, as well as the weakening of the economy (trade balance imbalance), and consequently, and abrupt devaluation of properties in the USA.
In this context, in 2005 Law 11.196 / 05, which became known as "Lei do Bem", creates the granting of tax incentives to legal entities that carry out research and development of technological innovation (MCTIC, 2019) to 2008, the pre-crisis period of 2008. In the case of the planted forest sector, this was an excellent trigger to perpetuate investments in R&D, since of the large companies in the planted forest sector, already had an R&D area before 2005. It should be noted here that this sector has a peculiarity regarding the temporal dynamics of research, for example, in the area of forest genetic improvement, that development of a clone of Eucalyptus lasts a day from 14 to 21 years to have a clone with proven productivity performance. In other words, the time needed to develop research in these areas is different from several other sectors of the economy, being the same in the long term, which requires strong and stable policies for the generation of research products for the planted forest sector.  Table 1 shows the high gross annual and total productivity gains, and the respective I / R ratios, showing how profitable the R&D activity is for the brazilian forestry sector. In view of these results, it can be inferred that the average return for each R$ 1.00 invested in R&D results in an average return of R$ 15.02 (1,502%). However, it is clear that the indexes ranged from R$ 1.28 to R$ 41.47 annually, obviously due to variations in gains in annual productivity, also due to climatic variations, clones etc. Similar results were found by FIGUEIREDO (2008) in which, for each R$ 1.00 invested, the return was R$ 13.67 for the orange crop in São Paulo. Other returns found by Araújo et al. (2002) for research in São Paulo agriculture and by Griliches (1975) for research in North American agriculture, of values between R$ 10 and R$ 12, and US$ 13, respectively. Evenson, Pray and Rosegrant (1999) found values between US$ 5 and US$ 6 for agricultural research in India, values much lower than those estimated here. It is worth mentioning that these authors had data about investments in research in the private sector, in addition to considering the effects of irrigation, which was not possible to incorporate in the present study. Consider that some investments in forestry R&D have very different payback times from others, such as: i) investments in forestry improvement require between 14 to 21 years to record performance clone performance and environmental adjustment; ii) investments in forest management experiments can take between 4 and 7 years to complete, for example, in relation to certain fertilizers with greater nutritional potential; iii) investments in operational development that take a few months to be concluded, for example on good mechanization practices and reduction of operating costs.
Making a comparative analysis in relation to the other types of low and high risk investments, it is clear that the average return on investment in R&D projects (1,502%) is much higher, such as, for example, those of IBOVESPA with an average annual return of 53.4% (p.y) (Santos et al., 2009); or even investments in fixed income with very low risk, but return only 6.5% (p.y.).
The variables in this study were correlated, and it was observed that marginal gains in wood productivity were positively correlated with investments in R&D (r = 0.43). This is very important is shows numerically that the investments in R&D made by forestry companies are giving It can be seen that over the period 2006 to 2013 the Brazilian forestry sector goes through a phase of intrinsic inflation in the sector above average inflation in Brazil, a systemic and gradual reduction in average wood prices and, therefore, a gradual and permanent reduction in profit margins of forest producers. On the other hand, there is the exporting industrial sector, which has a list of exporting companies associated with IBÁ and which is composed of a select list of 40 companies that operate in the foreign market and save exports of cellulose, paper, panels, sawn wood and floors laminates. The value exported by these companies in 2015 is approximately US$ 9 billion (IBÁ, 2016). According to Hersen et al. (2019) an exchange rate policy with sharp devaluation, initiated in 2012, has resulted in additional gains for the exporting timber forestry sector, in special industrialized products. However, these benefits are concentrated in a small number of large companies, with the possibility of further improving the performance of the forestry industry made for small producers, through opportunities also, or micro and small companies using the international market. In this sense, an antagonistic relationship is perceived between the prices and results of forest-based companies and the prices practiced in the wood market, which is the medium and small producer of wood that is being used, being used by other alternatives to increase the added value of wood so that the forestry business returns to execute in terms of profitability. On what: Index (I/R) = index that relates investments and revenues to R&D activities in forestry companies

Conclusions
In the period from 2006 to 2013 the Brazilian forestry sector goes through a phase of inflation intrinsic to the sector above the average inflation in Brazil, a systemic and gradual reduction in average wood prices and, therefore, a gradual and permanent reduction in the profit margins of forest producers.
Investments in R&D projects carried out by Brazilian Eucalyptus and Pinus forestry promoted gains in wood productivity in the period from 2006 to 2013.
Forest R&D is believed to be one of the most interesting business strategies for wood-producing companies to minimize the impacts of increased inflation, as well as the stable prices of commodities from forest products (cellulose, solid wood, briquettes, etc.) in national and international markets.
It can also be concluded with this work that investments in R&D projects translate economic returns for Brazilian forestry companies; on average, for each R$ 1.00 invested, companies obtain an economic return of R$ 15.02.