Balancing Risk and Learning Opportunities in Corporate Venture Capital Investments: Evidence from the Biopharmaceutical Industry

3.1 Research Design and Setting

To address the research question on the optimality of CVC venture portfolios in terms of risk-learning balance under scrutiny, we apply a qualitative, inductive field analysis. Such investigation is followed by the identification of several elements favoring the combination of risk exposures and learning benefits (Glaser and Strauss 1967; Eisenhardt 1989; Yin 1994). CVC deals represent our unit of analysis. More specifically, we collect secondary data on CVC transactions mostly drawn from Medtrack and Global Corporate Venturing. Conceptual variables are operationalized by measuring and mapping the empirical properties of collected data through scaling indicators (mainly at ordinal level) and subsequent scoring (or statistical) techniques.

Given the scarce empirical evidence on VC staging which permits to balance risk exposures and learning benefits, we follow the logic of grounded theory (Glaser and Strauss 1967; Yin 1994). Based on the observation of pattern matching across CVC transactions, theory is inductively advanced by proposing two propositions that, if empirically corroborated in future scholarly research, may suggest prescriptive ways of constructing CVC portfolios to accomplish (non)learning and/or risk mitigation (maximization) objectives. The adoption of a qualitative methodology is also consistent with the fact that the timing of CVC investment decision stages involves choices that are embedded in nature (Strauss and Corbin 1990; Langley 1999).

The research setting is the biopharmaceutical industry, which represents the confluence of the business activity of those large pharmaceutical firms active in the biotechnology sector, with specific geographic focus in U.S. and Europe. Such a setting is suitable for two main reasons. First, this industry has been identified as being characterized by radical innovation (Rothaermel 2001), making it an ideal context to analyze risk and VC staging (Li 2008). Second, knowledge innovation is dispersed within this industry, being shared by big pharmaceutical firms and new biotechnology firms (NBFs). This implies that innovation processes are driven by intense alliance (Rotheaermel and Deeds 2004; Cockburn 2004) and CVC activities (Zahra 1996; Henderson 2009). Although many studies have looked at the impact of CVC on innovation performance in the biopharmaceutical industry (Dushnitsky and Lenox 2005a), none has so far examined how the risk-learning balance may shape CVC investment portfolios.

3.2 Data Collection

Based on Medtrack ^[3] database, a unique dataset was constructed by collecting relevant information on global CVC deals mainly involving U.S. and European biotechnology firms in the 2003–2013 periods. More specifically, our focus is on CVC financing devoted to sponsoring new investments made by biotech companies to advance the internal development of outstanding compounds with a potential to become marketable drugs. Starting from a massive database of 5.138 transactions announced and/or completed by active CVC funds over the past 25 years (1989–2013) and available within Medtrack, three criteria were applied to make the selection of the final sample of our study. First, all deals for which CVC investors are undisclosed were eliminated so as to allow for the identification of all parties involved. Second, all deals involving biotech firms, for which the current R&D stage of the compound targeted by CVC financing could not be recognized, were disregarded. Third, out of the transactions fulfilling the above criteria, only those carried out by the major 26 CVC funds mostly operated by leading pharmaceutical or biotech firms were kept in. ^[4]

Unambiguously, the most active CVC funds operating in the biopharmaceutical industry were identified based on the ranking performed by Global Corporate Venturing (GCV) (June 2010). More specifically, only those deals involving biotech firms targeted by 22 funds (included in the top 30 CVC funds ranked by GCV) and other 4 funds (ranked among the top 60 CVC funds by GCV) operated by branded pharmaceutical companies, were actually selected. The list of CVC funds (with their relative GCV ranking) acting as investors in biotech companies included in our dataset is shown in Table 2.

The final sample of transactions completed by leading CVC funds targeting biotech companies and their relevant investments in a compound at current R&D stage is made of 260 deals spanning the period between 2003 and 2013.

3.3 Data Analysis

Our dataset corroborates the publicly available evidence (recalled in section 1) that CVC financing was a growing phenomenon in the biotech industry over the last decade, with a remarkable rise recorded from 2010 onward, which merits further investigation. Starting from the closing of only 11 deals in 2003, there was a steady increase of CVC activity worldwide – despite a slump in the number of CVC deals accomplished in 2008–2009 due to the severity of the financial crisis, followed by a sharp escalation in 2010 – that reached a peak in 2012 with the successful completion of 37 transactions (Figure 1).

Figure 1:

Number of CVC deals per year (2003–2013).

The degree of intensity of CVC activity in the biotech sector is also reflected in the total and average size of deal-making across the 2003–2013 period. In this respect, the total yearly amount of CVC financing steadily increases from $ 359 million to $ 1.139 million in the pre-financial crisis period (2003–2007) mimicking the increasing pattern in the number of transactions over the same years (Figure 2, Panel A). A decline in the total value of annual CVC deal-making is instead observed in the post-financial crisis (2008–2013) with a reversion to the initial bottom amount of $ 378 million in 2013. The pattern of average CVC deal values mirrors that observed at aggregate level. While the average market value of transactions carried out by CVC funds gradually increases in the pre-financial crisis period (2003–2007) rising from $ 32.6 million in 2003 to $ 35.6 million in 2007, the post-financial crisis period (2008–2013) is characterized by a sharp decline in the average deal value that drops to $ 21 million in 2013 (Figure 2, Panel B). If such a trend in the average value of deal-making is combined with the escalation of CVC activity in terms of number of transactions completed, it is fairly clear that in the past five years (since 2009) CVC funds show a preference to engage in an increasingly higher number of low capital-absorbing investments by investing less equity capital on average per single transaction. This corresponds to a risk-adverse, diversification approach to venture portfolio construction.

Figure 2:

Total and average CVC deal value per year ($ million) (2003–2013).

An in-depth analysis of the information on CVC transactions included in our dataset reveals some interesting peculiarities of the investment strategies pursued by the leading funds in the industry and of their target biotech firms.

First, the current R&D stage of the molecule being developed by each biotech firm, when targeted by the CVC fund, is examined with the aim of detecting the predominant CVC investment strategy. For this purpose, all molecules under development associated with each CVC transaction included in our dataset are grouped into six categories: Pre-Clinic Phase, Phase I, Phase II, Phase III, Drug Approval (e.g., compounds gradually developed into candidate drugs waiting to obtain approval from the national competent authority, such as the Food & Drug Administration – FDA in the United States), Market (e.g., approved drugs launched into the market). ^[5]

As shown in Figure 3, the number of CVC-funded molecules that are in early R&D stages (Pre-Clinic, Phase I) (75) at the time of the respective fund’s intervention is much lower relative to that of compounds positioned instead at later R&D stages (Phase II, Phase III, Drug Approval, Market) (185). More specifically, biotech companies conducting R&D activity on molecules at Phase II are mostly preferred as target investments by CVC funds, as 118 transactions (45% of the molecules underlying the deals included in our sample) are completed on such compounds. Molecules at Phase III, Drug Approval and Market are jointly considered as next best (67 transactions; 26% of the sample). Clearly, compounds that are at Phase II (or at a later stage) are more likely to become blockbuster drugs in the marketplace and should be thus regarded as less risky as compared to molecules still laying behind in their development process (e.g., those at research/discovery or Phase I stage). Consistently with Lo and Naraharisetti (2013), such evidence implies that over the 2003–2013 span CVC funds in the biopharmaceutical sector, through directing most of their capital at firms with a Phase II-compound under development, pursued a more risk-adverse approach to venture investing.

Figure 3:

Molecules per R&D stage financed by CVC (#; %) (2003–2013).

Second, the R&D targeting focus of CVC investment strategies is also analyzed across time (Figure 4). While in 2003–2005 CVC investors are not yet able to pursue a clear investment strategy in the biotech sector, starting from 2006 CVC funds adopt a dominant safe Phase II-focused investment strategy, which stands still throughout the remaining period. The construction of such principal risk-adverse portfolios is thus started by leading CVC funds since 2006 and mounts up over the most recent years.

Figure 4:

Number of molecules per year and R&D stage (2003–2013).

Third, most CVC investment strategies scrutinized in our sample are concentrated on U.S.-based biotech companies (195 investments), with U.K.-based ones being considered as next best targets (16 investments) (Figure 5). This indicates that leading CVC funds that are active in the biopharmaceutical industry tend to poorly leverage geographic diversification in their portfolio construction.

Figure 5:

Geographic distribution of CVC deals (2003–2013).

Fourth, the number of stages through which capital of CVC investors is typically infused into the biotech industry is also examined. CVC funds, and more in general VC investors, rarely finance target firms by providing all of the capital required by the venture upfront. Rather, investments are staged with periodic capital provided to the entrepreneurial firm subject to milestones being met (Gompers 1995). In the VC jargon, these subsequent stages of financing are named as “rounds” and labeled by a letter of the alphabet. For instance, the round of financing defined as “A” corresponds to the earliest event of capital provision, followed by rounds B, C, D and so forth. ^[6] The frequency of staged capital commitments may range between 6 months and 2 years. More frequent staging is impractical due to the legal and time costs of writing and negotiating contracts (transactions costs) and the monitoring activity being required to verify whether or not milestones have been met (monitoring costs). The stages of capital infusion are usually agreed upon between the CVC fund and the target firm based on the business plan projections. If the predetermined milestones cannot be met by the portfolio firm, it does not necessarily imply that the investment is ceased. It may simply mean that future capital provision will be revalued and the portfolio firm’s management must expect an increasing monitoring activity by the CVC investor.

One of the main explanations for the existence of VC funds – along with the risk-sharing (Amit, Glosten, and Muller 1990) and the skills learning (Chan, Siegel, and Thakor 1990) arguments – is the presence of agency problems (Amit, Brander, and Zott 1998). Due to agency costs, start-up firms without a track record and little or no collateral face extreme difficulty obtaining capital from banks. VC funds with expertise in carrying out due diligence for entrepreneurial start-ups with great asset specificity (industry-specific or firm-specific) are capable of mitigating problems of adverse selection and writing contracts to monitor investments and attenuating moral hazard and other agency problems that might adversely affect firm value (Cumming and Johan 2014). By staging capital injections, the CVC fund keeps the portfolio company on track with its committed business plan so as to minimize its own financial risk. In this sense, while staging capital infusion facilitates monitoring of portfolio companies and thereby reduces agency problems, it also forces the CVC investor to face a tradeoff between mitigating agency costs and incurring transactions costs (e.g., associated with issuances of new financing instruments) (Gompers 1995). Hence, in a world with both agency and transactions costs, a moderate degree of staging – that is, a clenched series (or number) of rounds of financing – is expected as optimal staging frequency would depend on the magnitude of monitoring and legal costs of contracting relative to the extent by which staged capital investment reduces agency problems. ^[7]

Consistently with Gompers and Lerner (1999), our evidence shows that CVC funds active in the biopharmaceutical sector are inclined to invest their own capital via use of up to three subsequent rounds (A, B, C) encompassing 175 transactions (67%). Interestingly, round B represents the mode of the rounds’ distribution (78 deals; 30%) implying that CVC funds of the above kind prefer to make follow-on, equity investments no more than twice into the same project (Figure 6). Agency problems (and related costs) are less pronounced in transactions that involve biotech companies as targets and pharmaceutical firms as parent entities of the fund given the proximity of their businesses, thus restraining the need for staged financing.

Figure 6:

Relationship between the amount of equity capital invested and the frequency of staged investments (number of rounds of financing) (2003–2013).

Moreover, there exists a positive relationship between the average amount of equity capital invested by the CVC fund and the timing of the round of financing (Figure 6). The earlier the round of financing, the lower the average amount of capital injected by the CVC investor into the target portfolio venture. In the VC world, the agency problems that staged investment practices seek to solve are motivated by the presence of significant information asymmetries between the investor and its target firm especially when the investment process is initiated. The size of staged VC capital injections is thus proportional to the degree of learning about the across-time, market potential of the target firm’s business plan that the investor gains as it progresses in the investment process. In the biopharmaceutical sector, as the CVC investor continues to financially support the R&D operations of the target firm, it can incrementally acquire new information on how likely the successful approval and subsequent commercialization of the molecule under development are. Indeed, the market launch of the candidate drug is the only source of cash flows for the target company, which in turn implies a high return at exit for the CVC investor. Our evidence confirms such industry practice as CVC investors tend to increase on average the amount of staged capital commitments as the number of rounds advances by infusing the uppermost amount of capital into the venture at the intermediate round C ($ 35 million). In so doing, CVC investors seek to minimize the agency problems posed by target biotech firms in their portfolios so as to keep their R&D operations on track toward the full development of the molecule that may become a marketable drug.

We finally inquire into the level of business maturity targeted by CVC investors. To do so, we refer to Maurer and Ebers (2006)s findings according to which a target portfolio venture in operation for no more than 4 years is considered as a startup. ^[8] Contrary to what CVC investors are expected to do, investment strategies pursued by funds operating in the biotech sector are – to some extent – aimed at targeting safer non-startups (141) rather than younger firms (119), whose R&D operations at earlier stages are affected by a higher risk of failure.

Conclusively, the descriptive examination of our sample provides a strong overall empirical evidence that in the 2003–2013 period CVC funds take a predominantly risk-adverse approach to constructing their venture portfolios by implementing risk mitigation strategies and engaging in the completion of less risky deals.

4.1 A Risk-Return Frontier for CVC Investments in the Biopharmaceutical Sector

Given the attention paid to risk in deal-making and portfolio construction by CVC funds, their investment strategies in the biopharmaceutical sector may be appropriately analyzed looking at the risk-return trade-off involved in the business venture selection process and, more in general, defined in the investment policy at the outset. Borrowing from finance and modern portfolio theory (Markowitz 1952, 1959), we thus develop a risk-return, efficient frontier for CVC investments that may serve as a tool for helping fund managers to locate new venture opportunities and measure their contributions to the overall portfolio risk and return.

To this end, we propose a measure for the attribution of a risk class (or score) to a stand-alone or portfolio-based CVC investment strategy. Such a risk score is defined by accounting for the relevant, current R&D phase of the compound being financed by the CVC investor by means of equity capital provision. More specifically, using Likert scaling (1932), ^[9] a notch of 1 is assigned to the commercialization stage (Market Launch phase), thus increasing by 1 point the notch attributable to any subsequent riskier stage so as to allot a notch of 6 (the highest) to the Pre-Clinical Phase, which should be considered as the riskiest stage in drug development. Indeed, a drug ready for commercial launch in the marketplace, having received the competent authority’s approval, is an ideal target for a potential, risk-adverse CVC investor due its very low intrinsic risk exposure. To the other extreme, a molecule at a Pre-Clinic stage is presumably highly risky because of its high probability of failing to succeed to next phases. As a result, each molecule in our sample is assigned a notch ranging from 1 to 6 that is inversely related to the underlying technical and/or market and thus financing risk assessed by the CVC investor.

A typical risk-return frontier is thus built by plotting the expected returns in the vertical axis and the associated units of riskiness in the horizontal axis for all investments collectively realized by CVC funds included in our sample. The expected return of each deal is derived by quantifying the cost of equity commonly required by VC investors and multiplying it by the size of the equity underwritten by the CVC fund. A rate range for the cost of equity used in the VC-backed biotech context is available from the empirical estimations provided by Cockburn (2004) and Cockburn and Lerner (2009). It follows that for molecules with a risk score of 4 (statistical mode) the cost of equity is based on an implied expected return of 21.5%. For molecules of lower riskiness (risk score of 3, 2 and 1) a decreasing cost of equity of 15.5%, 9.5% and 3.5%, respectively, is determined. For molecules of higher riskiness (risk score of 5 and 6) an increasing cost of equity of 27.5% and 33.5%, respectively, is estimated. The units of riskiness of each CVC deal of our sample in the horizontal axis are measured on the basis of the risk class (or score) derived above. The combination of expected returns and risk scores for all CVC market transactions available in our sample yields the mean-variance, efficient frontier (solid line) shown in Figure 7. If portfolios of venture investments, rather than single transactions, are considered, then the frontier arises from plotting aggregate expected returns and risk scores. Aggregate risk scores are derived from averaging out notches assigned to each specific deal.

Figure 7:

CVC Risk-Return Frontier.

Such a frontier represents a sort of CVC opportunity set along which a CVC fund can move to choose the optimal portfolio of investments as a function of its degree of risk aversion and its return expectations. Our proposed CVC efficient frontier may help CVC investors to optimally allocate their own equity capital based on the notion of an inherent risk-return trade-off.

4.2 A Typology of Option-Based Investment Strategies for CVC Funds

The main motivations driving companies to arrange CVC programs are the obtainment of attractive financial returns and the exploitation of strategic synergies and/or benefits resulting from the full integration (or a select knowledge spillover) of the new venture into the corporate entity that promotes the CVC activity (Chesbrough 2002). Among the benefits of CVC investments is the creation of new learning opportunities. However, as discussed in Sections 1 and 2, the extent to which a corporation may learn from new ventures acquired and nurtured by the captive CVC fund and the types of associated growth options subsequently exercisable at corporate level depends on the degree of overall risk exposure chosen by the latter for its investments through the pursuit of a portfolio diversification (low risk) or concentration (high risk) strategy.

To establish a relationship between learning propensity and risk exposure of CVC investors, we rely on our dataset to build two scores that, if properly combined, can be used to differentiate among CVC investment strategies. By analyzing the interplay between these two dimensions, we can identify a broad range of strategic options available to CVC investors and leading to a proactive venture portfolio management. A conceptual map, called as “CVC option map”, is next proposed to classify four typologies of CVC strategies (Figure 8).

Figure 8:

CVC option map.

Along the vertical axis is what we call the learning propensity score, which is calculated by multiplying – for each CVC fund included in our sample – the average deal value (in $ million) by the average number of rounds of financing executed for all portfolio transactions (where greater sub-scores are attributed to more frequent rounds such as A and B, that is the assigned sub-score decreases as incremental capital is injected into a portfolio venture by the CVC investor at the next round). On one hand, the average amount of equity capital provided by a CVC fund to its all portfolio companies reflects the degree of learning achieved by staging capital commitments and increasing them at each subsequent round across all transactions (in line with the evidence discussed in Section 3; see Figure 6). The higher the average value of transactions, the greater the level of knowledge acquired by the CVC investor on each of its portfolio firms. On the other hand, the average frequency of staged capital investments represents an adjustment to the average deal value to account for the fact that learning (from portfolio companies) must be moderated by the trade-off between the mitigation of agency problems and the disbursement of the costs of contracting (e.g., monitoring, negotiation). Consistently with Gompers (1995), while staging capital investments reduces the exposure to agency problems, it also involves frequent, costly transactions and monitoring activities. This favors a round frequency limited to the A-B-C sequence as opposed to a longer series (down to G). In this way, learning and monitoring (with associated contracting costs) are reconciled.

The learning propensity of a CVC investor is thus high when both the average deal value and the average number of financing rounds completed for all deals are high; intermediate, when the average number of financing rounds completed for all deals is high, but the average value of such transactions is low (or vice versa); low, when both quantities are low. A CVC fund showing a high learning propensity implies that it tends to maximize the learning process associated with the new venture investment by disbursing a large (and increasing) amount of equity capital through an optimally clenched series of financing rounds. When the timeline of rounds is more stretched and equity tickets are smaller, learning propensity diminishes. The values of the learning propensity score have a lower bound in 0 and are potentially limitless.

Along the horizontal axis is the Herfindahl–Hirschman Index (HHI) ^[10] measured on the basis of the molecule-picking (or selection) strategy and associated risk exposure pursued by CVC funds included in our sample. More specifically, the HHI, ranging from 0 to 1, quantifies the level of (portfolio) diversification (0) or concentration (1) for CVC investment portfolios.

As a result of the combination of the above two dimensions, CVC funds may be classified as (moderately) risk-lover or (moderately) risk-adverse when seeking to pursue four, option-based investment strategies. Starting from the most straightforward investment behavior, CVC funds positioned in the quadrant 4 of the CVC option map are less prone to learning from acquired ventures by tending to invest on average small equity tickets within a long series of rounds, with the purpose of constructing a concentrated venture portfolio. Those funds are focused and specialized organizations seeking to leverage their deal-making to allow parent firms to grow internally. For this reason, investments are concentrated into ventures associated with the discovery and development of molecules of the same type of those being developed by the parent firm. This entails less learning needs at corporate level, which translates into lower spillover efforts by the CVC investor. Target portfolio companies embed growth options exercisable by the parent firm to merely enhance and/or integrate its current business operations. The exercise of such growth options occurs via the pursuit of exploitation strategies mostly based on licensing agreements. The way the parent firm may benefit from its CVC initiatives is through learning associated with the establishment of licensing agreements with the portfolio company. CVC funds of this kind are labeled as “organic growth options seekers”. Due to a portfolio investment concentration on ventures mostly engaged in developing molecules of the same risk class through small equity capital infusion, such CVC funds are also classified as moderately risk-lover in terms of their overall risk profile.

CVC funds positioned in the quadrant 3 of the CVC option map are, similarly to those of quadrant 4, still engaged in building a concentrated venture portfolio but are more likely to be willing to learn by investing large equity tickets through a clenched series of rounds. Indeed, our sample-based empirical evidence shows that the value of market transactions completed by these types of CVC funds is above average and related financing occurs with less frequent staging. Their investment strategies are targeted at compounds featuring those development stages that are predominant in the parent firm’s R&D pipeline. The large size of the investment outlay allocated within an optimally compressed sequence of rounds maximizes learning, thereby facilitating subsequent knowledge absorption at the parent firm level through follow-on acquisitions of portfolio companies. Parent firms seek to capture growth options embedded in CVC-targeted new ventures that, given their high potential for value creation, require larger equity tickets paid in anticipation of an ensuing business integration. CVC funds of this kind may be defined as “external growth options seekers”. Blending a narrow, predominantly single-molecule focus of portfolio venture investments with a large value of the equity stake elevates the risk exposure of CVC initiatives, which may be then classified as risk-lover.

At the intersection of a low average deal value and a diversified venture portfolio investment strategy (quadrant 2), CVC funds are inclined to identify and potentially exercise (real) options to explore new markets and/or products by searching for breakthrough drugs or medical device technologies among several projects. Given their nature of unexplored territories, such CVC investment opportunities would have little potential to enhance the parent firm’s current business operations. However, if the CVC arm of a corporation succeeds in detecting – among the numerous ventures included in its portfolio – the innovative content of a distinct company that is capable of making (and bringing to approval) an important discovery advancing the pharmaceutical practice, the option to learn from such innovation may be exercised at parent level by transforming its future business operations with a novel and unique technology. In this context, learning is an exceptional event and the related process quite discontinuous. This renders learning propensity a less important attribute for these funds. The inherent riskiness of such explorative mode typically requires a small equity capital investment. CVC funds of this kind are classified as “breakthrough options seekers”. Risk-wise, these CVC investment strategies are of risk-adverse nature.

Finally, if a diversified approach to venture portfolio investing is combined with a great learning propensity, the strategic rationale of CVC initiatives situated in quadrant 1 is that of disbursing large amounts of equity capital across a broad spectrum of ventures in market transactions requiring an optimally compressed series of rounds so as to maximize the identification and potential, subsequent exercise of growth options available within the parent firm’s ecosystem. Parent corporations may be motivated to utilize CVC investments to stimulate the advancement of the ecosystem in which they operate. Such ecosystem is typically made of rival biotech companies engaged in research that is complementary to outstanding R&D pipelines conducted within the parent firm. A captive CVC program that leads to the purchase of equity stakes in these ventures by taking over their on-going research projects may give the parent firm an option to stimulate market demand for its existing or would-be drugs. Such CVC investment strategies may also be aimed at sponsoring competitive initiatives (e.g., technologies, devices, molecules) that would help parent firms to streamline their existing business operations. The learning potential of such CVC strategies is enormous as knowledge and skill spillovers from target ventures can be leveraged to enhance internal exploration at parent level in exchange for a moderate risk exposure due to portfolio diversification. CVC funds positioned in quadrant 1 of the CVC option map are thus classifiable as “ecosystem options seekers” and moderately risk-adverse.

The majority of CVC funds surveyed in our sample are positioned in the quadrant 1 of our conceptual map, which confirms the shift of CVC strategies toward a more risk-adverse approach to venture investing. This strategic change of CVC activity conducted in the biopharmaceutical sector may be explained by the recognition that corporate value creation may be improved if an explorative attitude is followed in venture scouting. Overall, the biopharmaceutical industry seems to be dominated by CVC programs featuring relatively high learning propensity regardless of risk attitudes (quadrants 1 and 3).

More interestingly, the conceptual option map constructed on the basis of our evidence shows that learning propensity and portfolio diversification are tied up by a curvilinear, U-shaped relationship. As CVC funds in the biopharmaceutical industry decrease the level of their portfolio diversification by heightening investment concentration, their learning propensity declines due to a lower number of available ventures that the parent entity can leverage to nurture its ecosystem and absorb knowledge. Their risk profile gets also exacerbated by more concentrated investments. After reaching a local minimum, learning propensity bounces back if risk-loving, focused portfolio investment strategies are pursued. R&D complementarity, massive equity investment and optimal fund staging help parent firms to target ventures that, if acquired, can accelerate value-creating knowledge transfer. Such a U-shaped relationship between learning propensity and the degree of portfolio diversification of CVC initiatives sponsored by pharmaceutical firms with the aim of targeting biotech ventures can be named as “CVC smile”. According to the above evidence, we suggest the following proposition:

Proposition 1: There is a curvilinear (U-shaped) relationship between CVC portfolio diversification and learning propensity.

Furthermore, learning-driven growth options embedded in CVC investment strategies mapped in our conceptual framework can be more generally grouped into two distinct categories. Learning options that are pursued by CVC funds of quadrants 3 and 4 to grow organically or externally may be classified as (real) options to expand parent business operations. Learning options associated with CVC initiatives comprised in quadrants 1 and 2 may be classified as (real) options to extend parent business operations. When exercising the former types of options, parent firms reach the objective of enlarging the dimension of their current business activities via CVC investment. Under this circumstance, learning is a purposeful activity aimed at reinforcing current R&D programs. When exercising the latter types of options, parent firms get access to unexplored markets or products so as to stretch their existing business lines by learning about novel technologies via CVC investment.

Depending on the amount of equity capital injections, the frequency of staged financing and the narrower or wider extent of portfolio investment (concentration vs. diversification), CVC funds may decide which risk/learning strategy to pursue by positioning themselves in one of the four quadrants of our CVC options map.

Portfolio concentration is the result of exploitation-based, focused investment strategies pursued to leverage CVC transactions to allow parent firms to grow internally or externally. Organic or external growth creates less valuable options as opposed to an explorative and diversified mode of investing. An investment in a breakthrough venture or a rival firm whose internalization could restructure the parent’s business may add more value to CVC portfolios (and, in turn, to corporate sponsors) than other types of assets. In the presence of portfolio diversification, growth and learning options are more numerous and of higher value impact.

Using our data on CVC deals in the biopharmaceutical sector, the combination of such dimensions as the level of portfolio diversification/concentration measured by the Herfindahl–Hirschman Index (HHI) and the value of the fund’s venture assets currently under management ^[11] provides empirical evidence on the negative relationship between these two factors and leads us to suggest the following further proposition:

Proposition 2: The level of a CVC portfolio concentration (and associated risk exposure) is negatively associated with the value of the investing firm's venture assets under management.

As shown in Figure 9, the higher the degree of portfolio concentration, the lower the value of the CVC firm’s venture asset portfolio.

Figure 9:

Relationship between portfolio concentration and CVC venture asset portfolio value.