Private Equity and Venture Capital in Brazil: An Analysis of its Evolution

This article focuses on the main idiosyncrasies of Private E quity and Venture Capital (PEVC) in Brazil and its evolution from 2004 to 2009. The m ain idiosyncrasies are the participation of limited partners in the investment process, absence of leveraged buyouts, shared control of the portfolio companies and use of special rights to compensate for the lack of full control. The main changes o bserved recently are the increase of private equity investments vis-à-vis v enture capital, in the efficiency of managers in the screening of proposals, in the use of pecial rights to compensate for the lack of control of the portfolio companie s, n the use of arbitration panels for conflict resolution, and in the participa tion of limited partners in the investment process.


Introduction
The literature on private equity and venture capital (PEVC) in Brazil is short, mostly due to the lack of data. Checa et al. (2001) present the history of the PEVC industry in Brazil in its early years (prior to 2000). Mariz & Savoia (2005) compare Brazil and the USA in order to exploit the challenges for the PEVC in Brazil. Supported by interviews with five PEVC organizations, Pavani (2003) also describe the critical factors for the development of the industry. Based on questionnaires answered by CEOs of portfolio companies (PC), Botelho et al. (2003) study value addition by general partners. Based on a survey of general partners (GP) in the period when the Brazilian IPO market was inactive, Ribeiro & Almeida (2005) found that the exit strategy influences the entire investment cycle, and that trade sale was the preferred means of exit. Finally, Minardi et al. (2013) study the long term price performance of Brazilian IPOs and find that PEVC-sponsored IPOs outperforms Non-PEVC-sponsored ones.
De Carvalho et al. (2006) were the first to provide systematic data on the characteristics and practices of the Brazilian PEVC industry. They obtained information from all managing organizations with office in the country. ABDI (2011) was the second systematic data collection, updating De Carvalho et al. (2006). Ribeiro & De Carvalho (2008) discuss how PEVC adjusted to the Brazilian environment and present a comparison with PEVC in the US. The main similarities were: industry composed mostly of independent organizations, institutional investors are the main source of capital, both commitments and portfolio are regionally concentrated, and IT is the most invested industry. The main Brazilian idiosyncrasies are the near absence of leveraged buyouts and concentration of investments in late stages.
This article focuses on the main changes that occurred in the Brazilian PEVC between 2004 and 2009, the two year for which there are systematic data on the PEVC industry. Our analysis is based on several sources, but mainly on De Carvalho et al. (2006) and ABDI (2011). Our study differs from the previous ones because it discusses the evolution of the PEVC industry and relates it to the changes that occurred in the macroeconomic environment. Furthermore, we point out and try to explain some idiosyncrasies such as the participation of investors in the investment process and absence of leverage buyouts. The main facts that we explore are: the evolution of aggregate commitments that increased more than six fold between 2004 and 2009, the increase of PE vis-à-vis VC, from 33% to 47% of the PCs, investment concentrated in some few sectors but the most invested ones have been changing, managers are becoming more efficient in the screening of proposals, absence of buyouts, use of special rights to compensate for the lack of full control of the PCs, increased use of arbitration panels for settlement of conflicts, and increasing participation of limited partners (LP) in the investment process. The remaining of this article is organized as follows: Section 2 presents the macroeconomic environment in which the PEVC industry has developed. Section 3 describes the recent evolution of the industry. Section 4 focuses on the organization of the PEVC funds. Finally, Section 5 discusses the investment process.

Macroeconomic environment
The steep rise of the Brazilian PEVC industry in the beginning of the XXI century can be related to a myriad of changes occurring at the macroeconomic level. The main changes were: improvement of the macroeconomic conditions along with the achievement of investment grade; resumption of economic growth; blooming of the stock market (rise of Novo Mercado); the emergence of the pension fund industry; and improved income distribution and poverty reduction. Below we discuss each one of these elements. Macroeconomic Stability: during the 80's and part of 90's, Brazil experienced two-digit monthly inflation rates. The macroeconomic stability achieved in the beginning of this century (inflation rates of about 5% per year) is consequence of a process initiated in 1994 with the Real Plan. It reduced inflation rate from about 50% to less than 1% per month. Along this process, stabilization was threatened on several occasions, e.g., during the currency crises in Mexico (1995), Asia (1997), Asia (1998), Russia (1998), Brazil (1999), Argentina (2001, and over the years 2001 and 2002 with the foresight of a leftist candidate being elected for president. In all these opportunities, the commitment to monetary stability was reaffirmed, especially in the early years of the leftist government of President Lula da Silva (2003 and2004). As consequence, inflation has been maintained at rates near 5% per year (Table 1). An important milestone was the investment grade achieved in 2008. 1 It caused foreign investments to reach the record of US$ 34.5 billion in 2007 and US$ 45 billion in 2008, more than twice of its previous values; Resumption of economic growth: a consequence of macroeconomic stability was the resumption of economic growth. The Brazilian economy had been stagnant for several decades. The average GDP growth was 3% in the 80's, 1.6% in the 90 's, and 2% over 1999's, and 2% over -2003's, and 2% over . Between 2004's, and 2% over and 2010 the average growth rate was 4.5% per year in spite of the global financial crisis. The resumption of economic growth was important to enhance the deal flow; Development of capital markets: the maturity of the Brazilian stock market was achieved mainly with the creation of BM&FBovespa 2 premium listings. According to De Carvalho & Pennacchi (2011), the little protection offered to minority shareholders was a major obstacle for the development of capital market in Brazil (use of non-voting shares, low level of disclosure and other bad governance practices). In 2000, Bovespa created three new market segments: Novo Mercado and Corporate Governance Levels I and II. To be listed on one of these segments the company needs to commit to a set of corporate governance practices.  (Table 1). It should be noted that the IPO market remained active even after the start of the global financial crisis in 2008, although much more selective; Emergence of the pension funds industry: by 2001 the existing retirement system was fully State operated. Active workers made contributions to the State based on their wages; the state used these contributions to pay the pensions of retirees; and pensions were based on the wages that retirees received at the time of retirement. Only some few large State-owned enterprises had pension funds. The aging of the population and increase of life expectancy led the State model of retirement to bankruptcy. As consequence, in 2001 the model had to be reformed: retirement age went up, and a cap for pensions was established at about US$ 1,400 a month. This led Brazilians to seek supplementary retirement through private pension funds. Since then this industry has been growing rapidly. From 2001 to 2009, the value of pension funds assets more than quadrupled from US$ 72 to US$ 324 billion (Abrapp, 2011); and Improved income distribution and poverty reduction: income concentration and exclusion of a large segment of the population from the consumption market was a feature of the Brazilian economy up the 90's. The proportion of the population below the poverty line in 1999 was 35%. The Gini index that measures income concentration was 0.59 (compare this to Uruguay, 0.45; Russia, 0.37; South Korea, 0.32; and USA, 0.41). Since the mid 90's, the Brazilian government created and intensified social policies for income redistribution and poverty alleviation. The total value of benefits increased from US$ 151 million in 2004 to US$ 632 million in 2009. 3 This caused the proportion of the population below the poverty line to fall to 21.4% in 2009 and Gini index, to 0.54. One should also mention the policies for the extension of credit to the low-income population. For example: payroll loans, 4 created in 2004, reached US$ 8 billion in January 2010. As result of resumption of economic growth and better income distribution, 32 million people moved from the lower class to middle class CPS-FGV (2010). The increasing purchase power of low-income population is creating demand for several products targeted at this segment. Many PEVC have been focused on low income, e.g., large scale low fee universities, building and construction, and low-income consumption goods.

Recent evolution of PEVC industry
The evolution of aggregate commitments (Table 1) shows a sharp rise of PEVC in Brazil. Between 1999 and 2004 commitments grew at 9% per year from US$ 3.7 to US$ 5.6 billion. More recently, from 2004 to 2009, the growth rate near 50% per year, reaching US$ 36.1 billion in 2009. Taken as proportion of GDP, the growth represents an increase from 0.63% in 1999 to 2.33% in 2009. Compared to countries where PEVC is more developed such as US (3.7% of GDP) and UK (4.7%), there is still room for continued growth. One also observes that the concentration of the capital commitment decreased from 2004 to 2009, although the average fund size increased (Table 2). In 2004, the 15 largest organizations were responsible for 76% of the total commitments. These organizations in average managed US$ 283 million. The other 56 existing organizations in average managed only US$ 26 million. In 2009, the 15 largest organizations managed 63% of the commitments. These organizations in average were responsible for US$ 1,519 million. The other 125 management organizations in average were responsible for $ 106 million.  Carvalho et al. (2006) and ABDI (2011).

Table 3
Stage of portfolio companies at the first finance round Distribution of portfolio companies in December 2004 and 2009 according to the stage where it received the first finance round, seed capital: pre-operational stage. start-up: the structuring stage of the business when the products aren't sold; expansion: expansion of the activities of a company that already sells its products; late stage: the company who has a stable growth rate and positive cash flow. Other Stages: includes Acquisition finance (capital to acquire other companies), mezzanine (stable companies with growth potential), management buyout / in (contribution for the acquisition of control), bridge finance (capital prior to IPO), turnaround (capital for companies with operational and / or financial difficulties) and PIPE (private investment in public equity) investment in companies already The rise in the number of portfolio companies is another evidence of the industry growth (Table 3). It increased from 306 PC in 2004 to 502 in 2009. There was also a change in the focus of investments, with an increase on private equity vis-à-vis venture capital. The proportion of PCs that received their first investment in the VC stage dropped from 67% to 47%, while the proportion of PE investments went from 33% to 53%. The relative increase in PE may be related to the increase commitments and arise of exits through IPOs, as conjectured by Jeng & Wells (2000). Surprisingly, investments in seed capital and start-up that require relatively small capital investment fell from 11.8% and 23.5% to 7.6% and 13.2% respectively.
In terms of industrial sector, investments were also very concentrated, even though the concentration fell. Table 4 shows that between 2004 and 2009 the participation of five main sectors increased from 51% to 57% of the PCs. The IT sector remained as the most invested, but its share fell from 30% to 20%. The sectors with highest growth were: Construction (from 3% to 14%) and Energy and Fuels (from 2% to 11%). The growth of energy sector was related to the growth in such market. Brazil will be the seventh largest energy market in 2030 (EYT-FGV, 2007). To meet the growing demand for energy, it will be necessary that the supply of energy increase at 3.3% per year over the next three decades. Furthermore, Brazil is one of the leading countries in the development of clean energy. The growth of investments in building and construction was related to factors such as: expansion of mortgage lending (from US$ 1.6 billion in 2002 to US$ 14 billion in 2007 according to YET-FGV (2008)); increase in the income of the poorest; and chronic housing deficit (7.8 million homes in 2005, according to YET-FGV (2008)).
Investments were geographically concentrated in some few Brazilian states (De Carvalho et al., 2006, ABDI, 2011. The South and Southeast regions concentrated the majority of the portfolio with 278 PCs (91% of the portfolio) in 2004 and 441 companies (92% of the PCs for which there is information) in 2009. Investments were concentrated even in some states inside these regions, e.g., São Paulo State had 44% of the portfolio in 2004 and 57% in 2009.  The means of exiting investments also changed over time.

Legal structure
In the US, limited partnerships (LP) are the most appropriate legal structure for PEVC (Sahlman, 1990). Its main advantage is that profits are not taxed at the partnership level. Each investor collects tax on profits at their own tax rate. Thus, pension funds, major investors in PEVC, do not collect tax on their PEVC investments. In Brazil there is no legal structure such as limited partnership. PEVC funds have used diverse legal structures. De Carvalho et al. (2006) show that the holding company was the most widely used fund structure. Limited partnerships structured offshore were also commonly used. The holding company model in Brazil, however, has two important shortcomings for PEVC: profits are taxed at the holding company level and there are limits to the appropriation of tax shields generated by losses on unsuccessful investments. To circumvent these limitations, Comissão de Valores Mobiliários 5 (CVM) established investment funds -FIP and FMIEE 6 -that allow complete appropriation of tax shields and investors to collect income tax at their own tax rates. From 2004 to 2009 the proportion of investment funds structured as FIP or FMIEE increased from 33% to 42%. 7 In the opposite direction, funds structured as holding companies decreased from 20% to 13%. The proportion of funds structured as LP offshore remained stable: 30% in 200430% in and 26% in 200930% in (ABDI, 2011. One of CVM requirements for FIP and FMIEE funds is the use of arbitration panels for settlement of conflicts. With the increased use of such funds, from 2004 to 2009, the proportion of funds that use arbitration panels raised from 19% to 76% (De Carvalho et al., 2006)

Funds life
The distribution of funds' life did not change significantly from 2004 to 2009 De Carvalho et al. (2006), ABDI (2011). Most vehicles had economic life between nine and 10 years (34% of funds in 2004 and 32% in 2009). Funds with duration between seven and eight years came next with 25% in 2004and 27% in 2009. Between 2004and 2009 appeared funds with duration of less than 5 years (4% in 2009). Covenants foreseeing possible life extension were very common. The most frequent extension period was two years (43% in 2004 and 41% in 2009). The use of other extension periods increased: one year extension rose from 3% to 13%; and 3 years, from 3% to 21%. Funds with defined life and without extension period increased from 13% to 21%.

Limited partners intervention in the fund management
The participation of limited partners in the fund management is an important peculiarity of PEVC in Brazil. De Carvalho et al. (2012) argue that in principle such participation could be functional: when the leverage buyout is the predominant business, GPs must obtain the approval of creditors for each one of the deals (Axelson et al., 2009).
However, in economies where there is no readily available long-term credit, as in Brazil (Souza Sobrinho, 2003), leverage buyouts are not possible. As consequence, there are no external agents to referee for GPs investment decision. Therefore, oversight from LPs could be appropriate (at least in principle).
LPs in Brazil have acting in the investment process either directly or through investment committees. Their interference varies significantly depending on the type of majority required for a decision in the committee (simple majority, qualified majority or unanimity). Table 6 shows that LPs control over the decision to invest in a new company increased. From 2004 to 2009 the proportion of funds with investors committee went from 55% to 65%. The proportion of funds that require qualified majority increased from 18% to 30%, becoming the most used structure. The proportion of blind pools (structure in which the LPs have no active participation) fell from 28% to 15%. The proportion of pledge funds (structure in which each individual LPs have complete control over participating or not on each proposed investment) remained stable at near 6%.  Carvalho et al. (2006) andABDI (2011). LPs participation in the investment process sometimes goes beyond the screening of proposals. It is quite common that LPs be in direct contact with PCs and even take part in the exit decision. From 2004 to 2009 the proportion of managing organizations that promote at least one meeting a year between LPs and PCs increased from 53% to near 65% (ABDI, 2011). In 2004, GPs had full control over the exit decision in 38% of the funds, the investment committee in 31%, and GPs and LPs decided together in 38%. This picture is slightly different when one takes commitments as reference; the funds for which GPs had full control represented 65% of the commitments, investment committees over 17%, and GPs and LPs together over 16%. According to De Carvalho et al. (2006), this result suggests that GPs of large funds have more control over in the investment process than those managing small funds. 8 To our knowledge, De Carvalho et al. (2012) is the only study that empirically examines the effect of LPs participation in the investment process. They found that funds with investment committee, when compared to those without it, present inferior performance.

Compensation
GPs compensation usually has two components: management fee and carried interest. The basis for the management fee usually differs across investment and post-investment periods (De Carvalho et al., 2006, ABDI, 2011. During the investment period, the most common basis is the commitment (55% of the funds in 2004 and 63% in 2009). For the postinvestment period there was a growing use of market value of investments (from 29% to 45%) and the cost of investments (from 12% to 27%). The use of commitments in the post-investment period declined from 38% to 12% of the funds. In 2004, the most frequently used management fee was either 2% or 2.5% of the basis (used in 41% of the funds). Also in 2004, 62% of the funds used carried interest of 20%. The most common model (used in 39% of the funds) provided carried interest of 20% and management fee of either 2% or 2.5%. 9

Affiliation of managing organizations
Similarly to countries in which PEVC is well established, independent organizations are emerging as the predominant form of managing organization, while the number of organizations affiliated to financial institutions or to the State has declined. The proportion of independent organizations increased from 63% in 2004 to 85% in 2009; financial institution affiliates declined from 28% to 12%; and State affiliates decreased from 3% to 1%. The picture is similar when we one considers capital commitments: independent organizations increased from 54% to 80%; financial institution affiliates declined from 38% to 16%; and State affiliates maintained their share at approximately 3%. Interesting, one should note the near absence of corporate ventures: four funds in 2004 and only two in 2009.

Origin of managing organizations
Managing organizations with Brazilian origin were predominant. The number of affiliates of European institutions increased, while that of North-American ones decreased. The participation of domestic organizations remained stable between 2004 and 2009 at nearly 75% of the organizations and 60% of commitments. The participation of North-American organizations decreased from 14% to 2% (from 10 to 3). However, in terms of commitments the fall was not so sharp: from 31% to 18%. Organizations from Europe increased from 6% to 16% (from 3% to 9% in terms of commitments).

Deal flow and screening
In line with the rise of the industry, the number of investment proposals analyzed doubled between 2004 and 2009, from 840 to 1681 per year. However, the proportion of deals that went through due diligence decreased. In 2004, 17% of proposals that were analyzed went through due diligence; in 2009, only 5.5%. The proportion of companies that had undergone due diligence and were invested increased significantly from 25% to 54%. These results suggest that GPs passed through a learning process and improved their screening skills. Apparently, they became more demanding when analyzing proposals, thereby reducing the possibility of rejection during due diligence. Another evidence of improved screening ability is the fall in the proportion of managing organizations that evaluated the same proposal more than once from 65% to 30% (De Carvalho et al., 2006, ABDI, 2011.

Closeness to portfolio companies
Two of the most important activities of GPs during the investing process are monitoring and value addition. Both activities require closeness between GPs and PCs. The closeness increased over time. An important indicator of these activities is the frequency with which boards of directors (BD) of PCs meet. From 2004 to 2009 the percentage of management organizations for which the BDs of PCs meet at least once a month increased from 47% to 67% (De Carvalho et al., 2006, ABDI, 2011. Similarly, the percentage of organizations for which the interval between meetings is greater than two months fell from 31% to 19%. Another indicator of closeness is the number of hours in meetings per month that GPs spend with their PCs. In 2004, 35% of the management organizations (responsible for 52% of the aggregate portfolio) dedicated 5 to 10 hours a month in meetings and telephone conversations with each PC; 31% of the organizations (responsible for 24% of the portfolio) spent 10 to 30 hours per month. Surprisingly, 18% of the organizations spent less than 5 hours per month. 10

PEVC without equity control
A peculiarity of the Brazilian economy is the unavailability of longterm credit. Commercial banks generally supply only short-term loans with terms not exceeding one year (Souza-Sobrinho, 2003). The only significant source of long-term loans is Banco Nacional de Desenvolvimento Economico e Social (BNDES). However, their loans are limited in value and restricted to priority projects such as, infrastructure, incentives for import and export, acquisition of equipment and machinery produced in the country. Because of this, leveraged buyouts are not characteristic of PEVC in Brazil, as in the US. Concomitantly, the need for diversification of PEVC funds limits the amount that can be invested in one single company. As consequence, GPs are unable to acquire the equity control of large companies. Alternatively they have to share the control. In 2004, PEVC investors were part in the controlling group of only 115 of a total of 325 deals. PEVC investors with a minority stake occurred in 26.2% of the deals. The absence of full control was usually made up by the acquisition of special rights such as veto power on key matters. In 2004, minority participation with veto power occurred in 38.5% of the deals (De Carvalho et al., 2006). ABDI (2011) reports that in 2009 61% of the deals embedded some special right to compensate for the lack of equity control. 11 The most common ones were: the right of veto in cases of mergers and acquisitions, 95% of the deals; on the investment budget, 91%; on the annual budget, 86%; on indebtedness, 86%; right to subscribe new capital, 46%; and drag along, 15%. Another frequently foreseen right is that of preferred return on invested capital. In 2009 about 23% of the deals had this covenant. The most frequent terms were: invested capital plus minimum return in case of tradesale, 79%; recovery of invested capital in the event of liquidation, 36%; and dividend income, 36%.