Volume 27 Issue 2 Article 3 June 2025 Startup Accelerator Returns: J Curve or L Curve? A Comparative Startup Accelerator Returns: J Curve or L Curve? A Comparative Performance Analysis Between a Venture Accelerator and Early- Performance Analysis Between a Venture Accelerator and Early- Stage Venture Capital Stage Venture Capital Aleš Pustovrh University of Ljubljana, School of Economics and Business, Ljubljana, Slovenia, ales.pustovrh@ef.uni-lj.si Follow this and additional works at: https://www.ebrjournal.net/home Part of the Entrepreneurial and Small Business Operations Commons, Finance Commons, and the Finance and Financial Management Commons Recommended Citation Recommended Citation Pustovrh, A. (2025). Startup Accelerator Returns: J Curve or L Curve? A Comparative Performance Analysis Between a Venture Accelerator and Early-Stage Venture Capital. Economic and Business Review, 27(2), 115-129. https://doi.org/10.15458/2335-4216.1355 This Original Article is brought to you for free and open access by Economic and Business Review. It has been accepted for inclusion in Economic and Business Review by an authorized editor of Economic and Business Review. ORIGINAL ARTICLE Startup Accelerator Returns: J Curve or L Curve? A Comparative Performance Analysis Between a Venture Accelerator and Early-Stage Venture Capital AlešPustovrh University of Ljubljana, School of Economics and Business, Ljubljana, Slovenia Abstract This document analyses the protability of investments in venture accelerators compared to early-stage venture capital funds. Using a case study of a single fund manager operating both investment types, it tracks the Total Value to Paid-In (TVPI) ratio over 6 years. The early-stage venture capital investments showed a positive trend, exceeding a TVPI of 1, indicating protability driven by company survival rates, external funding attraction, and growth. Conversely, the accelerator investments underperformed, with a TVPI consistently below 1, suggesting a loss for investors. This raises questions about the long-term viability of the accelerator model, potentially resulting in an L curve rather than the expected J curve of returns. While the accelerator’s performance could still improve if the few successful companies signicantly outperform the underperforming majority, this reliance on a small number of successes represents an inherently higher risk for investors. Future research should incorporate broader datasets and consider various market dynamics to generalize the ndings, utilizing panel data across different geographies and industries. Keywords: Venture capital, Accelerators, Investment performance, Startup funding, Return on investment JEL classication: G24, L26, M13 1 Introduction and the theoretical background on startup accelerators B usiness accelerators help prospective startups develop initial business solutions, dene and identify their customer segments, and provide re- sources, including capital and employees (Cohen & Hochberg, 2014). Drawing from resource dependence theory and open innovation paradigm, we propose that a business accelerator can be seen in the function of resource provision to growing companies along with the monitoring and control role that it can have in the ecosystem. Within the entrepreneurship ecosystems a spe- cial role is reserved for the “startup factories”— business or venture accelerators (Miller & Bound, 2011). A venture accelerator (or startup accelerator) is a xed-term, cohort-based programme that provides early-stage companies with seed funding, intensive mentorship, educational resources, and access to a network of investors and industry experts. Accord- ing to Cohen and Hochberg (2014), accelerators are designed to rapidly accelerate the development of startups through rigorous selection processes, con- centrated mentorship, and a structured curriculum that culminates in a “demo day” where startups present their progress to potential investors. This model not only supplies critical early-stage capital and guidance but also enhances a startup’s credibility within the broader entrepreneurial ecosystem. In the last decade a new generation of incubation models, i.e. the seed accelerator programme, emerged as a response to the old models, mostly focusing on pro- viding ofce spaces and in-house businesses (Bruneel et al., 2012). The main features of the accelerator programmes are preseed investment (usually in exchange for eq- uity), time-limited support, with an emphasis on Received 11 March 2025; accepted 7 May 2025. Available online 10 June 2025 E-mail address: ales.pustovrh@ef.uni-lj.si (A. Pustovrh). https://doi.org/10.15458/2335-4216.1355 2335-4216/© 2025 School of Economics and Business University of Ljubljana. This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/). 116 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 intensive mentoring, networking and educational programme for cohorts or classes of startups (Miller & Bound, 2011). The programme usually lasts from three to six months and concludes with a demo day when incubated startups present their project to a large number of investors (Kim & Wagman, 2014). Key elements of startup accelerators thus include (Pustovrh, 2019): • seed capital & equity stakes: Programmes usually offer a modest investment in exchange for a small equity share. • mentorship and training: Founders receive guid- ance from seasoned entrepreneurs, industry ex- perts, and sometimes former accelerator alumni. • cohort-based learning: Entrepreneurs benet from peer-to-peer interaction, shared learning experiences, and collaborative problem solving. • access to networks: Accelerators connect startups with venture capitalists, corporate partners, and other ecosystem players, often opening doors that would be difcult to access independently. Initial research related to this topic provides de- tailed description of accelerators, their functioning and programmes (Isabelle, 2013; Miller & Bound, 2011), their distinction to incubators and other sup- port mechanisms, and the importance of the various aspects of the programmes to the success of their incu- batees (Cohen & Hochberg, 2014). It nds the primary distinguishing features of accelerators in the limited duration of the programmes, for classes of startups who enter and graduate together. There is an emerging consensus that accelerators have a role in the success of their startups. According to a recent comprehensive multinational study of more than 8000 startups supported by 408 accelerators in 176 countries (Assenova & Amit, 2024), accelerated startups were more likely to raise venture capital (VC), raised more capital in the 1st year after graduating from these programmes, and planned to raise more capital afterwards, on average, over the next year. Accelerated startups also generated more revenue, hired more full-time employees, and paid more in wages to their employees, on average—indicating that they scaled faster than their peers (Assenova & Amit, 2024). They have an estimated 23% higher survival rate than their peers (Regmi et al., 2015). Other ndings suggest an overall positive impact on accelerated startups in terms of acquisition of new knowledge (Battistella et al., 2017; Wise & Valliere, 2014), startup valuation (Kim & Wagman, 2014; Smith et al., 2016), entrepreneurial ori- entation (Hayter et al., 2018; Stayton & Mangematin, 2019), and ability to receive subsequent funding (Radojevich-Kelley & Hoffman, 2012). Evidence highlights the importance of pursuing higher efciency over the life cycle of a startup (Balboni et al., 2019). Accelerators aid startups regardless of their pre-acceleration growth status (i.e., growing, stagnat- ing, or declining). Furthermore, nongrowing startups selected for acceleration exhibit better chances of achieving growth postacceleration compared to already growing counterparts (Tekic et al., 2024). The performance of venture accelerators is inu- enced by several key variables, as highlighted in selected recent studies. Cánovas-Saiz et al. (2020) emphasize the importance of portfolio size, startup survival rates, and the number of employees in ac- celerated rms, noting that these factors positively impact the median funding received by startups. Additionally, the longevity and geographical loca- tion of accelerators, particularly those in the U.S., are shown to enhance startup survival rates. Moroz et al. (2024) lend support to the positive impact that venture accelerators may have across multiple lev- els of observation—for example, accelerated startups’ levels of survival, investment funding (timing, lev- els, rounds, and speed), growth measures (employ- ees, revenues, valuation), exit pathways (acquisition, IPO, sale), and a wide range of intangible measures (new/faster product development, entrepreneurial efcacy, cognitive bias effects, and key milestones). Their structured review underscores the necessity of a comprehensive evaluation approach that considers both quantitative and qualitative metrics to accu- rately gauge accelerator performance. Academic critiques of the venture accelerator model emphasize several limitations inherent in its design and reporting practices. For instance, the reliance on a cohort-based structure has been criticized for introducing survivorship bias, as success stories tend to dominate public narratives while failures remain underreported (Cohen & Hochberg, 2014). This focus on the most successful outcomes can mask the true risk prole and variability of performance among startups in an accelerator, thereby providing an overly optimistic picture of the model’s efcacy. Moreover, the intensive networking and shared experience within cohorts may inadvertently promote groupthink, reducing the diversity of innovative approaches and potentially stiing alternative strategies that could lead to breakthrough success (Smith et al., 2016). Other scholars have drawn attention to the issue of asymmetric bargaining power between accelera- tors and startups. Critics argue that accelerators, by virtue of their inuential position and network ad- vantages, may secure disproportionately favourable equity terms from startups—potentially compromis- ing the long-term value creation for the entrepreneurs ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 117 involved (Hallen et al., 2020). In addition, research suggests that opaque reporting practices in accel- erator performance further complicate an unbiased evaluation of returns and risks, calling for more rig- orous, transparent reporting of startup valuations (Charoontham & Amornpetchkul, 2023). These cri- tiques highlight the need for further research to rene the accelerator model and develop more comprehen- sive metrics for assessing its impact within the VC ecosystem (Leitão et al., 2022). Another gap in the literature is the research on ac- celerators in peripheral entrepreneurial ecosystems. The establishment of Y Combinator acceleration programme (in the USA) pointed to the active role of business accelerators in an entrepreneurial ecosystem, and they have grown rapidly in the U.S. ever since, followed by a trend replication in Europe (Miller & Bound, 2011). Business accelerators have become an interesting subject of scholarly research only later, with scholars investigating the demand and supply of business incubation services across different incubator generations (Bruneel et al., 2012), exploring the accelerators’ incentives to choose a portfolio size and disclose information about partici- pating ventures (Kim & Wagman, 2014) or focusing on a specic type of business accelerators, for example, corporate accelerators (Kohler, 2016). One of the rst studies focusing exclusively on European business ac- celerators (and excluding other technology business incubation mechanisms, such as technology parks, incubators, and innovation centres) was provided by scholars (Pauwels et al., 2016) who proposed a typol- ogy of accelerators, that is, the “ecosystem builder,” the “deal-ow maker,” and the “welfare stimulator.” Although the study delivered a rich overview on how accelerators operate as a new-generation incubation model and how they differ from existing incubation mechanisms, the research focused on the three “leading accelerator regions” in Europe: London, Paris, and Berlin, which indicates an opportunity and need for future research in other countries (Mian et al., 2016). However, as the main body of research evidence is based on data from developed economies of the U.S. and Western Europe, there may be some discrepan- cies in the evolution of the business accelerators in less developed markets (Uhm et al., 2018). More im- portantly, the role that business accelerators have in the startup ecosystem might be different in emerging startup ecosystems than in the more developed and resource-abundant environments. Research exploring how an accelerator can help and speed up the devel- opment of entrepreneurship ecosystems, how it can help less developed entrepreneurship regions, and how policy makers can help towards development is lagging behind—although not completely absent either (Pustovrh et al., 2020). Another topic not well researched is the commercial success of the acceler- ators and the factors inuencing their success—for example, their geographic or industry focus. Even a simple general question—“Are investments into venture accelerators protable?”—lacks a clear an- swer, with the public generally pointing towards a string of successful companies accelerated by Y Com- binator but failing to research its overall protability or factors inuencing its rate of return on invested capital. To gain more insights on whether venture ac- celerators represent a viable investment opportunity for private investors seeking their required rate of return—such as early-stage VC funds—more empir- ical research is required. There is a clear gap in the academic literature evaluating the nancial perfor- mance of accelerators across different markets—and even in the markets with the most developed accelera- tor ecosystems. While the importance of public policy interventions for support of accelerators is widely available (European Commission & Organisation for Economic Co-operation and Development, 2019), its effects on the accelerator performance—its rate of return—are not sufciently researched. However, this question is important. If they are protable, there is a clear market solution to provid- ing acceleration services to startups. Essentially, they would represent another segment of the VC market— the preseed- and seed-stage-focused VC. If, conversely, venture accelerators consistently do not deliver market-rate returns, this suggests the presence of market failures or externalities that call for consideration for public intervention, such as subsidies, grants, or other forms of support aimed at fostering innovation and entrepreneurial activity. Such interventions can be justied if they address information asymmetries, coordination problems, or positive externalities associated with startup accelera- tion (Grilli et al., 2018). The “ecosystem builders” role could be an example of this (Pauwels et al., 2016). However, these externalities will have to be evalu- ated and measured to determine the scope and modes of public intervention and support of accelerators. Even if accelerators would not provide positive re- turns on investments by themselves, there might still be appetite for corporate investments into corporate accelerators—essentially supplementing nancial re- turns for strategic benets that such organizations would contribute to large corporations (Kohler, 2016; Kupp et al., 2017; Weiblen & Chesbrough, 2015). This piece of research is thus focused on answering the following hypothesis: Is there empirical evidence that a venture accelerator model can deliver a positive rate of return on its 118 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 investments that is comparable to returns in early-stage VC funds? 2 Expected returns of VC funds To answer our hypothesis, it makes sense to treat accelerators as a specic form of early-stage VC fund as, like VC funds, they make investments into a set of early-stage startup companies and obtain an equity stake in them. VC is a form of private equity nancing provided by investors to early-stage, high-growth companies in exchange for an equity stake, with the aim of fuelling their rapid expansion and innovation (Hayes, 2024). VC investments are often specialized according to the stage of business development of the investee company: • preseed: This is the earliest stage of business de- velopment, when the startup founders try to turn an idea into a concrete business plan. They may enrol in a business accelerator to secure early funding and mentorship. • seed funding: This is the point where a new busi- ness seeks to launch its rst product. Since there are no revenue streams yet, the company will need a VC fund’s or accelerator’s funding to fund all of its operations. • early-stage funding: Once a business has devel- oped a product, it will need more capital to ramp up production and sales before it can become self- funding. The business will then need one or more funding rounds, typically denoted incrementally as Series A, Series B, and so forth. Preseed and seed stage startups are typically funded by business angels or accelerators while more traditional (and usually larger) investments that are required by early-stage startups are provided by more traditional VC funds. This distinction is not clear and many early-stage VC funds also run their own acceler- ators and vice versa. Y Combinator, for example, pri- marily identies itself as a startup accelerator. Its core programme is a three-month, intensive cohort-based experience that provides early-stage companies with seed funding, mentorship, and access to a robust net- work of entrepreneurs and investors. However, it has also expanded its investment activities—for example, through the YC Continuity Fund, which supports its accelerated companies in later growth stages. VC has garnered considerable attention amidst the upsurge of entrepreneurship, particularly in the United States, as noted by Kaplan and Schoar (2005). Their paper, “Private Equity Performance: Returns, Persistence, and Capital Flows,” provides one of the most cited empirical analyses of private equity performance, including VC and buyout funds. The study primarily examines fund-level returns using measures related to Multiple on Invested Capital (MOIC) and Total Value to Paid-In Capital (TVPI). MOIC is a performance metric that measures the total return on an investment by comparing the sum of its current residual value plus any distributions received to the total capital invested. Essentially, it indicates how many times over the initial investment has been returned, though it does not account for the time value of money. TVPI is a comprehensive measure of a VC fund’s performance. It is calculated by dividing the sum of the current residual value of the investment portfolio and the distributions re- turned to investors by the total paid-in capital. This ratio reects both realized and unrealized returns, providing a full picture of a fund’s overall return on invested capital. Both TVPI and MOIC measure fund performance using the total value (realized plus unre- alized) of private equity investments. The difference is in the denominator. MOIC divides the total value of the investment or fund by the total invested capital, while TVPI divides it by the paid-in capital (Albers- Schoenberg & Zeisberger, 2019; Moreira, 2023). The J curve is a graphical representation illustrating the typical return trajectory of VC funds over time. In the rst few years, these funds often experience negative returns due to management fees, operational expenses, and the early stages of capital deployment. This early decline is followed by a period where returns begin to improve as portfolio companies mature, gain traction, and potentially achieve protable exits, resulting in positive returns for investors. This progression forms a curve resembling the letter “J” (Meyer & Mathonet, 2005). The J-curve effect highlights the importance of long-term investment horizons in VC, as signicant returns are generally realized only after several years of nurturing and developing portfolio companies. Empirical evidence shows that VC funds typically require between 5 to 7 years to reach an inection point where cumulative returns turn positive (Kaplan & Schoar, 2005). Several factors contribute to the J-curve effect in VC funds (Grabenwarter & Weidig, 2005): 1. early-stage expenses: At the outset, funds incur management fees and operational costs before any signicant investment gains materialize, leading to initial negative returns. 2. investment maturation: Investments in startups require time to develop, scale, and reach liq- uidity events. During this period, the fund’s valuations may remain static or even decline un- til successful exits occur. ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 119 Fig. 1. Fund standard J curve. Source: Meyer and Mathonet (2005) Fig. 2. Signal jamming problem masking the underlying quality of a venture. Source: Hellmann et al. (2024) 3. value impairments and write-offs: Some port- folio companies may underperform or fail, re- sulting in impairments and write-offs that can further depress early returns. It is important to note that the presence of a J curve does not necessarily indicate poor performance; rather, it reects the inherent life cycle of venture investments. As successful portfolio companies achieve exits, the gains can offset early losses, leading to the upward slope of the J curve. Un- derstanding this pattern helps investors set realistic expectations about the timing of returns in VC investments. However, it can also send investors mixed signals about the quality of their portfolio or reveal differences between portfolio managers (Chan et al., 2020). The model presented in Fig. 1 of Hellmann et al. (2024) illustrates the classic J-curve phenomenon in VC investing by mapping out cash ow trajectories over time. In the gure, high-quality ventures are de- picted with two distinct J curves—one corresponding to a short-term strategy and another to a long-term strategy—while a low-quality venture, employing a short-term approach, shows a shallower curve. The model presented in Fig. 2 captures a “signal jamming” problem: although a high-quality venture ideally prefers a long-term investment strategy to unlock its full growth potential, doing so may mask its underlying quality by making its initial negative cash ows appear similar to those of a low-quality venture. Consequently, the model highlights the trade-off 120 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 Fig. 3. Signal jamming problem in venture accelerators. entrepreneurs face between enhancing short-term cash ows to secure follow-on nancing and preserv- ing long-term cash ow potential, with investors’ loss tolerance—shaped by their exit opportunities— playing a pivotal role in determining the optimal strategy. In order to evaluate whether accelerators with their portfolio of investments in very-early-stage startup companies represent a viable investment, a clear gap in the literature has been identied—there are very few empirical studies analysing the rate of return of early-stage VC funds and no studies on the rate of return on investments into venture accelerators. An intuitive model on the expected J-curve cash ows for startups of different quality in Fig. 1 provides a good framework also for analysing the differences between early-stage VC funds and venture accelerators. While a well-performing portfolio of early-stage startups (in either long-term or short-term quality) is expected to follow a J curve and eventually show positive rates of returns, a portfolio of startups in the accelerator could potentially remain in the negative returns, indicating lower quality (or higher risk) of such startup investments—essentially following an L curve instead of the expected J curve. On the other hand, it might also simply take longer time for such a portfolio in an accelerator to become protable, indicating a longer-term strategy towards the expected (higher) rates of return of a portfolio of startups in the venture accelerator as the investments by the accelerator were implemented in an earlier development stage of startups. Investors cannot know their portfolio’s underlying quality as its initial negative cash ows appear similar to those of a low-quality venture. As seen in Fig. 3, it might simply need more time to mature. This signal jamming problem on the quality of different portfolios has not been researched empirically so far in early-stage VC funds or venture accelerators. 3 Empirical evidence on the J-Curve returns in VC VC returns have been the subject of extensive em- pirical and theoretical inquiry, with several studies highlighting the interplay of risk, timing, and di- versication in shaping fund performance. Sahlman (1990) laid the foundation by contrasting venture- capital organizations with large public corporations and leveraged buyout rms, emphasizing the unique agency issues and contractual solutions that char- acterize VC structures. This early work underscores how the inherent high risk in early-stage investments is managed through carefully designed operating procedures and investor relationships. Building on this framework, Prencipe (2017) presents an analysis based on a hand-collected dataset of approximately 3600 VC investments backed by the European Investment Fund (EIF) spanning from 1996 to 2015. The study reveals that VC returns are not static; rather, they show sensitivity to economic cycles and follow a power-law distribution, suggesting that a small number of investments generate outsized returns. Moreover, the work highlights that diversication strategies, enabled by the heterogeneity of deals, help to mitigate risk and lower the correlation of VC returns with broader ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 121 Table 1. Expected total value to paid-in capital (TVPI) for venture capital funds. Fund quartile Expected TVPI (Year 6) Top quartile 1.5–2.0 Median fund 1.2–1.4 Bottom quartile 1.0 Source: Extrapolated estimations based on Kaplan and Schoar (2005). asset classes. Notably, rm experience—particularly in later-generation funds—appears to bolster performance, although this might also reect the stringent screening of rst-time teams by the EIF. Complementary insights are offered by Weidig and Mathonet (2004), who compare risk proles across investment vehicles, nding that direct VC invest- ments bear a roughly 30% probability of total capital loss, whereas VC funds and funds of funds benet from diversication effects that substantially reduce this risk. Swildens and Yee (2017) further rene this understanding by proposing a risk-and-return ma- trix specic to VC, which situates these investments within the broader private equity landscape. Recent market analyses, such as those reported by Clemens (2024) on net cash ow J curves, corroborate the delayed yet robust return prole of VC funds, un- derscoring the importance of temporal dynamics in performance assessment. In their study, Kaplan and Schoar (2005) nd that, on average, private equity funds (including VC funds) have an MOIC close to 1.8 times the invested capital at their maturity (after about 10 years). For VC funds specically, the TVPI (a measure that includes both realized and unrealized returns) tends to be above 1.5 at maturity in top-quartile funds and closer to 1.0 in median funds. Since private equity and VC funds generally follow a J-curve pattern, in Year 6 and extrapolating on these results, a typical fund would likely have a TVPI between 1.2 and 1.4, depending on fund quality. Given that many funds distribute capital primarily after Year 6, the implied MOIC and TVPI in Year 6 would be modestly above 1.0 for the average fund and signicantly higher for top-tier funds. Kaplan and Schoar (2005) thus imply that by Year 6, the TVPI for early-stage VC funds typically ranges from 1.2 to 1.5, with top-tier funds exceeding 1.5. Lower-performing funds might still hover around 1.0, indicating limited appreciation at that stage. The expected TVPI values for VC funds are presented in Table 1. These results align with the broader J-curve effect, where many VC funds are still in the value creation phase in Year 6. This will be the benchmark used for our empirical analysis. 4 Methodology and data collection 4.1 Sample and setting To evaluate whether investments into venture accel- erators are protable, a set of data on returns on two portfolios was collected: a portfolio of investments into startups that were accepted into an accelerator and another portfolio of investments into startup companies invested from an early-stage VC fund. Both portfolios were managed by the same fund man- agement company with clearly dened investment criteria. The fund management company operates in Central and Eastern Europe and is sector-agnostic but geographical-region-specic. It is a suitable case for analysis as it manages both the accelerator and the early-stage VC fund, making investments into both from two separated compartments and distinct investment strategies but with the same investment committee. Individually, both the fund and the accelerator structures resemble those of typical accelerators and early-stage VC rms. Combined, the single fund management for both funds is quite an exceptional, although not unique, example and is thus specically suitable for addressing the research ques- tion. With full access to their performance data on a quarterly basis, this makes it a good case study for our analysis. Preseed and seed-stage investments of up to EUR 200,000 were invested into startups that did not achieve a product-market t yet (and sometimes lacked any revenue whatsoever). These companies were accepted into an accelerator programme. Each 3-monthly programme would provide standard ac- celeration services in addition to the investment: workshops and seminars or webinars, mentor ofce hours, access to support networks (other startups, alumni, potential local customers, perks, and other organizations in the local startup ecosystem as well as a concluding demo day). In the rst 5 years, 10 such programmes were implemented and a total of 113 in- vestments (and EUR 7.4 million) were implemented from the accelerator. In parallel, early-stage investment was limited to startups that had already achieved product–market t and had at least some initial and growing revenue and potential for future growth. Investments into such companies were larger, starting at EUR 250,000, and could be followed by additional investments if these companies continued or even accelerated their growth. The largest amount invested into a single company from this early-stage VC fund was EUR 1.5 million. During the fund’s 5-year investment period, a total of 32 investments (and EUR 29.4 million) were implemented from this early-stage VC fund. 122 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 Both portfolios were industry-agnostic and not lim- ited to digital/software startups only—although such startups represented a clear majority of investees. Due to the large number of investments, startups from al- most all industries were supported, except for those from a standard list of restricted sectors. 1 It is important to note that some startup compa- nies that have been accepted into the accelerator were also able to secure the early-stage VC invest- ment, usually at EUR 250,000 or slightly more. There were 21 companies that were able to receive both accelerator and early-stage investments. With these companies, selection bias could be an issue if they were systematically different from those receiving only one type of investment. This represents an op- portunity for future research. In the current analysis, companies that received both investments had the same estimated valuation as their value used in the accelerator compartment was always equal to their value in the early-stage VC compartment. 4.2 Portfolio valuation methodology and the valuation measures used To analyse the differences in the value progression of both portfolios, an objective valuation methodol- ogy was developed for all startups invested from both compartments—the early-stage VC and the accelera- tor compartment. Its aim is to establish a reasonable valuation for each portfolio company and thus for both portfolios. Since private company valuations are inherently complex, an accepted methodological ap- proach was developed that reduces subjectivity by applying weighted external measures to decide an as- sessed valuation consistently across reporting periods and compartments (startups in the accelerator or in the early-stage VC fund). The method incorporates three valuation bases: • revenue multiple • cost basis adjusted for short runway and liquida- tion preferences • external funding rounds All portfolio companies were categorized into three groups based on investment age and growth trajec- tory. Specically, three main indicators from startup quarterly reports provided key observations used to categorize them: their revenue growth, their cost growth with an estimation of their runway (projected number of months before they run out of currently held cash reserves), and external funding rounds used as a proxy for market-based transaction setting the valuation of the company: a) 1st-year investments: Investments made within 12 months of the reporting date are typically held at cost unless extreme negative circumstances arise, or an external funding round has occurred. Follow-on investments by the fund extend this period, but valuation remains at cost. b) growing investments: Investments beyond the 12-month window that demonstrate growth— based on individual targets but never below 10% month-on-month—were valued using a weighted combination of revenue multiples, cost basis, and external funding rounds. The most important reason to increase the value of the company in the portfolio was always in the external round. If an external funding round has occurred, the postmoney valuation was incorporated, provided the investment rm played a minority role or did not take part. c) stalled investments: Investments not meeting expected growth projections are subject to impairment. Discounts are applied based on the potential recoverability of the business or underlying assets, often resulting in signicant write-downs. For example, if a company from the accelerator was not able to raise an external round and its growth and development consis- tently lagged behind its own plans and investor expectations, and the company had short run- way, the value of such a company was impaired by 25% each quarter, resulting in a total impair- ment within 1 year. In general, 25% valuation impairment per quarter was the norm for under- performing companies, but not all of them were fully impaired if the investment compartment had certain preferential rights such as a liquidity preference. Each startup company was evaluated using these measures every quarter. Based on revenue, cost and potential external funding rounds reported by each portfolio startup company, these data were cross- checked with the annual nancial statements to as- sure their reliability. For external funding rounds, the companies were contractually obliged to report all such transactions and their terms. These data were cross-checked with industry benchmarks for their respective industries in order to validate that post- money valuations were in line with them. 1 A typical example is available at Guidelines on the EIF Restricted Sectors. ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 123 Table 2. Quarterly Total Value to Paid-In (TVPI) for the early-stage venture capital (VC) compartment and the accelerator (ACC) compartment. TVPI (early-stage TVPI (ACC No. of companies No. of companies Time VC compartment) compartment) invested (VC) invested (ACC) Q3 2019 0.61 0.12 5 3 Q4 2019 0.87 0.79 10 22 Q1 2020 0.63 0.52 10 25 Q2 2020 0.7 0.61 14 35 Q3 2020 0.71 0.69 21 54 Q4 2020 0.77 0.7 25 65 Q1 2021 0.75 0.63 33 72 Q2 2021 0.69 0.61 39 77 Q3 2021 0.8 0.52 44 80 Q4 2021 0.94 0.58 45 87 Q1 2022 0.89 0.64 46 92 Q2 2022 0.99 0.69 50 93 Q3 2022 0.93 0.69 50 100 Q4 2022 1.03 0.67 50 105 Q1 2023 1.06 0.6 50 105 Q2 2023 1.28 0.67 52 106 Q3 2023 1.34 0.61 52 106 Q4 2023 1.33 0.64 53 113 Q1 2024 1.34 0.66 53 113 Q2 2024 1.51 0.66 53 113 Q3 2024 1.53 0.64 53 113 Q4 2024 1.8 0.59 53 113 Note. There were 21 companies that were able to receive both accelerator and early-stage VC investments. 4.3 Analytical technique The methodology assigns weightings to different valuation approaches depending on the company’s circumstances. The revenue multiple approach ap- plies industry-standard benchmarks: 2 • 5x annual recurring revenue (ARR) for hardware sales (nonrecurring revenue) • 10x ARR for service-based models (nonrecurring revenue) • 20x ARR for recurring service revenues • 30x ARR for recurring service revenues with strong customer retention The cost basis valuation is adjusted based on nancial runway, with discounts applied for short cash reserves and further adjusted for liquidation preferences if applicable. The nal valuation assigns weightings based on the relative importance of each method. For instance, a recent external round may receive a higher weighting (up to 100%) in valuation, while a company struggling to secure funding may see greater emphasis placed on cost basis adjustments. Obtaining a valuation of each startup company in both portfolios each quarter was the basis for computing the net asset value of the whole portfolio in the accelerator and the early-stage VC fund separately. This methodology ensures a consistent and struc- tured, data-driven approach to VC valuation in both the accelerator and early-stage VC fund, mitigating subjectivity while reecting the nancial and opera- tional realities of portfolio companies. 5 Results and analysis The valuation of both portfolios was conducted on a quarterly basis and for both the accelerator and early- stage VC fund at the same time. The resulting startup valuations, together with information on the values of capital calls and management fees, were used to calculate both MOIC and TVPI. Other performance metrics (Internal Rate of Return [IRR], Distributed to Paid-In [DPI]) were also considered, but DPI was re- jected as most of the value gains in startups were not realized yet and DPI measures only realized gains, while IRR is used less often and thus not good for comparing results with other studies. For benchmark analysis, the resulting TVPI was used for comparison between the accelerator and early-stage VC fund in their 6th year of operations. The results are presented in Table 2 and Fig. 4. For the early-stage VC portfolio, J-curve results are clearly observable. While the TVPI was initially relatively low at the very beginning of both operations due to establishment costs being paid 2 Grinda (2020, 2021); Miller (2022); Ostin (2021); Sarath (2021); microcap.co. (2021); Wilhelm (2021). 124 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 Fig. 4. Total Value to Paid-In (TVPI) comparison between accelerator and early-stage venture capital (VC) fund in their 6th year of operations. Fig. 5. The observed J-curve returns for the early-stage venture capital portfolio and L-curve returns for the venture accelerator after 6 years of operation. in Q1, the Q2 TVPI result for the early-stage VC portfolio was 0.87. In the same period, the accelerator portfolio’s TVPI is comparable at 0.79. As predicted by the J curve, TVPI for both portfolios falls to much lower, and in both cases actually the lowest, values—0.63 for the early-stage VC portfolio and 0.52 for the accelerator portfolio. However, the two portfolio returns started to differ from this point forward. The TVPI for the early-stage VC portfolio began to increase, mostly due to the rela- tively high survival rate of invested companies as well as relatively good results for some of them—some companies in this portfolio started raising money from external funding sources (typically other VC ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 125 funds). The TVPI for this segment reached 0.8 by Year 3 and exceeded 1 in the 4th year (14th quarter of op- eration). As a select few companies in this portfolio embarked on accelerated growth and their valuations reached much higher values, this more than replaced the write-down for those companies that underper- formed. By Q2 in Year 6 (22nd quarter of operation), this portfolio’s TVPI reached 1.8 and was clearly on path to increasing further. On the other hand, the accelerator compartment did not achieve this growth in TVPI. In fact, its TVPI at the end of the observed period (22nd quarter of op- eration) was 0.59, lower than in Year 1 (4th quarter). While the TVPI in the accelerator portfolio recorded both increases and decreases in the observed period, it was generally stagnant and below 1, indicating that the accelerator compartment was returning less than it was invested into—thus losing money for its in- vestors. In this case, the answer to our hypothesis—Is there empirical evidence that a venture accelerator model can deliver a positive rate of return on its investments that is comparable to returns in early-stage VC funds?— is clear: the venture accelerator portfolio was not protable in the 6th year of operation. The venture accelerator portfolio’s TVPI value remained well be- low 1. On the other hand, the early-stage VC portfolio investments were protable in the same time period, showing a clear growth trend in increasing protabil- ity after the initial decrease—thus following a clear J curve. The early-stage VC portfolio started making positive returns after 4 years and was close to reach- ing a TVPI of 2 at the end of its 6th year. As the stagnant trend of TVPI-measured returns for the accelerator was very clear, it is also possibly the most likely outcome that the accelerator will continue losing money throughout its 10-year investment pe- riod. As presented in Fig. 5, its J curve would thus turn into an L curve. It might be possible that the answer to our research question on the protability of invest- ments into an accelerator will remain negative—at least the trend of the rst 22 quarters of this accelera- tor would imply that. 6 Discussion and implications, limitations, and future research avenues As the typical length of a VC fund is 10 years and both studied portfolios were in their 6th year of oper- ation, the results of the case study will only become clear in the future. There are some factors that might suggest that the TVPI of the venture accelerator might increase in the future. The most obvious one is that the few successful accelerator investments need more time to start rapid growth—due to their very early stage when accepted into the accelerator, they simply require more time to show clear results. And their re- sults need to be extremely good as a large proportion of accelerated startups eventually do not bring any returns on investment—so the remaining few need to make much larger returns, and these take more time. In our case study, out of 112 venture accelerator investments, 21 have received a follow-up investment from the early-stage VC fund from the same fund manager, but only 3 have increased their value by a factor of 5 or more (but less than 10)—which is not enough to show positive results for the whole venture accelerator’s portfolio. Their numbers might still increase, but most of the accelerated companies have already had their value impaired, some of them fully. While this is expected and a normal result of the high-risk prole of such early-stage investments, it nevertheless requires the few successful startups to make so much larger returns. For example, if the exit success rate of accelerated startups is only 1%, the increase in the value of those startups that make a re- turn must reach 100 times the initial investment value just to return the initial investment into the whole portfolio. Such “lottery-ticket winners” are rare, and there is a clear possibility that the accelerator portfo- lio will struggle to reach even positive values in the remaining 4 years. There is also another possibility. The reason for stag- nant TVPI in the accelerator compartment is that only a few companies were able to embark on a path of fast growth, raising external funding rounds in the process. A large majority of the portfolio was unable to achieve that and would either become a small, break-even lifestyle business or would not survive. In both cases, the valuation methodology would require the valuation of such companies to be signicantly or completely impaired. The combined decrease in the value of those companies was signicant enough to outweigh the increase of the few companies in the portfolio whose values were increasing. However, as the investment period of 5 years has passed, the impairment of nonperforming investments will even- tually lead to full realization of losses, leaving only a handful of companies whose value is expected to keep increasing. If the combined value of these few selected companies’ valuations keeps increasing, this trend will eventually overcome the valuation impair- ments that would cease due to the total elimination of the nonperforming part of the portfolio. One can even imagine a hypothetical extreme case where the valuation of a single accelerator-invested company (a “unicorn”) could be worth much more than the combined valuation of the fully impaired portfolio of all other invested companies. This is the case in the Y Combinator, which has successfully created several 126 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 such “unicorn” companies. But not many other accel- erators can show such success. In such a case, the accelerator’s returns would still follow the J curve, but at a much slower rate than in the early-stage VC fund. Their returns would also be more dependent on a small number of successful startups, their expectations for prolonged fast growth, and their ability to attract external investors to sup- port it. However, even in such cases this dependence on a lower number of successful startups represents a greater risk than in the early-stage VC fund. That means that this case study represents a clear, if incomplete, answer to the research question. It pro- vides evidence that a venture accelerator does not deliver a positive rate of return for its investors, while an early-stage VC fund does provide such a positive return for its investors. 6.1 Practical implications There are a few clear lessons for venture accel- erators resulting from this study. Investors should understand that this is highly risky investment opportunity that depends on few selected successful startups and a large majority of startups whose value will be partially or fully impaired. An accelerator should thus make investments accordingly—into bold new ideas implementing disruptive innovations and potentially creating and capturing a lot of value. Individual startup investments should most likely estimate the chances of such an investment making a return of 100 times the initially invested capital—even 10-times returns on a single investment will not be sufcient to make the investments into the whole portfolio protable. It should also consider steps towards spreading the same investment potential to a larger portfolio of startups to increase the chances of supporting a startup with the huge growth potential required. Portfolio size has been shown by other studies (Cánovas-Saiz et al., 2020) to be important to individual accelerated startup results—our study shows that larger cohorts are also positive for the investors into the accelerators. Another outcome of this case study provides a clear rationale for public policy intervention. If pri- vate investors cannot make a required rate of return investing into assets as risky as the accelerator’s portfolio, public intervention is justied—especially if other positive externalities can be observed. This brings other risks, mostly connected to the selection of startups, where not every startup company has an opportunity to create value—so how can acceler- ator managers be incentivized to select those startups that are high-risk but have at least a potential for high return and not simply invest public money into startups that realistically cannot expect to make re- turns on investment? In such a case, an investment from an accelerator would effectively become a grant and as such even represent state aid. It seems that a private–public-partnership could provide a solution, for example by mixing private investment capacity and capabilities with public investments. There are good examples of public organizations and institu- tions acting as lead or anchor investors into private accelerators but requiring additional private money to be invested as well, including from the accelera- tor management. While such arrangements are quite common in VC funds, they are much less popular in the nancing of venture accelerators. 6.2 Theoretical implications This study advances the existing literature on ven- ture accelerator nancing and VC performance by juxtaposing empirical results and comparing them to a nuanced theoretical framework. It shows that while J-curve returns are observed in early-stage VC, such returns in a venture accelerator might not be real- istic and might resemble an L curve—and thus not provide the positive returns that venture accelerators would require to make investments into such organi- zations. This would also favour public intervention to support such organizations in case clear positive ex- ternalities would be obtainable. While prior research has acknowledged the J-curve phenomenon—where startups experience initial losses before achieving protability—our work delves deeper into the strate- gic trade-offs between short-term protability and long-term growth. This study bridges a critical empir- ical gap in the literature by linking investor loss toler- ance with startup performance outcomes in two dif- ferent but comparable investment portfolios, offering valuable insights for entrepreneurs and policymakers aiming to foster innovation and economic growth. 6.3 Future research avenues The primary limitation of this case study method is its reliance on data from a single fund manager run- ning both an early-stage VC fund and an accelerator, positioned in a specic context of the emerging nan- cial systems of Central and Eastern Europe. While this approach allows for in-depth analysis, it raises questions about the external validity of the ndings. To increase robustness, extending the analysis be- yond 6 years would be crucial to observe later-stage performance dynamics. In order to increase the va- lidity of results, an additional interview with another, older venture accelerator in the same region was con- ducted; it made its rst investments more than 10 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 127 years ago and has also made investments into 10 co- horts through a period of 5 years. While a slightly dif- ferent valuation methodology was used, that acceler- ator’s portfolio results were quite similar to our obser- vation and its portfolio’s rate of return on its invest- ments remained below 1 even after 10 years. While no more than an indication, this observation hints that the L-curve portfolio returns might remain the same even through a longer period than the one observed in our case. However, in order to verify these expected results, we plan to repeat our analysis in 4 years’ time. To generalize these results from a single case study, other case studies involving a broader set of similar funds would be necessary, focusing on actual perfor- mance metrics such as returns on investments. The specic geographic and industry contexts of this fund manager could limit the applicability of the results. Different regions or sectors may show distinct market dynamics and funding landscapes, particularly in later stages of startup growth. These variations could signicantly inuence startup valuations and overall fund performance, thereby affecting the generalizability of the study’s conclusions. Including data from multiple accelerators and early-stage VC funds to rule out idiosyncratic factors related to the specic fund manager studied would greatly increase the robustness of the empirical results observed in our case study. Future research could also benet from a formal econometric approach controlling for other factors (e.g., industry, stage of investment, macroeconomic conditions). Adopting panel data methods to analyse a broader array of accelerators across different ge- ographies and industries would also be desirable and would offer additional insights. This approach would enable researchers to capture longitudinal changes and assess how varying economic, regulatory, and market conditions inuence accelerator performance and VC returns over time. By including diverse ac- celerators in such a panel, scholars could conduct comparative analyses that reveal underlying patterns and distinct operational efciencies across different environments, ultimately leading to a more nuanced understanding of the VC ecosystem. Finally, the role of lead partners as investors into VC funds and venture accelerators, particularly funds of funds, presents a crucial avenue for future inquiry. These entities often control access to privately collected data that are not publicly available, which makes them key stakeholders in understanding the full spectrum of investment outcomes. Moreover, public and multilateral organizations that act as anchor investors in many accelerators and VC funds could further enrich the analysis. Their involvement may help standardize data collection and offer insights into the external factors that contribute to the generalizability of research ndings, thereby strengthening the overall validity of future studies. References Albers-Schoenberg, A., & Zeisberger, C. (2019). Measuring private equity fund performance. INSEAD. https://www.insead.edu/ sites/default/les/assets/dept/centres/gpei/docs/ Measuring_PE_Fund-Performance-2019.pdf Assenova, V ., & Amit, R. (2024). Poised for growth: Exploring the relationship between accelerator program design and startup performance. Strategic Management Journal, 45(6), 1029–1060. https://doi.org/10.1002/SMJ.3581 Balboni, B., Bortoluzzi, G., Pugliese, R., & Tracogna, A. (2019). Busi- ness model evolution, contextual ambidexterity and the growth performance of high-tech start-ups. Journal of Business Research, 99, 115–124. https://doi.org/https://doi.org/10.1016/j.jbusres .2019.02.029 Battistella, C., De Toni, A. F., & Pessot, E. (2017). Open acceler- ators for start-ups success: A case study. European Journal of Innovation Management, 20(1), 80–111. https://doi.org/10.1108/ EJIM-10-2015-0113 Bruneel, J., Ratinho, T., Clarysse, B., & Groen, A. (2012). The evo- lution of business incubators: Comparing demand and supply of business incubation services across different incubator gen- erations. Technovation, 32(2), 110–121. https://doi.org/10.1016/ j.technovation.2011.11.003 Cánovas-Saiz, L., March-Chordà, I., & Yagüe-Perales, R. M. (2020). New evidence on accelerator performance based on fund- ing and location. European Journal of Management and Business Economics, 29(3), 217–234. https://doi.org/10.1108/EJMBE-10 -2017-0029 Chan, C. S. R., Patel, P . C., & Phan, P . H. (2020). Do differences among accelerators explain differences in the performance of member ventures? Evidence from 117 accelerators in 22 coun- tries. Strategic Entrepreneurship Journal, 14(2), 224–239. https:// doi.org/10.1002/SEJ.1351 Charoontham, K., & Amornpetchkul, T. (2023). Reputational im- pact on startup accelerator’s information disclosure and per- formance. Economics of Innovation and New Technology, 32(2), 250–274. https://doi.org/10.1080/10438599.2021.1910030 Clemens, J. (2024, April 27). Examining the net cash ow J-curves of recent PE fund vintages. PitchBook. https://pitchbook.com/ newsletter/examining-the-net-cash-ow-j-curves-of-recent-pe -fund-vintages Cohen, S., & Hochberg, Y. V . (2014). Accelerating startups: The seed accelerator phenomenon. SSRN. https://doi.org/10.2139/ ssrn.2418000 European Commission & Organisation for Economic Co-operation and Development. (2019). Policy brief on incubators and accel- erators that support inclusive entrepreneurship. https://doi.org/ 10.2767/092345 Grabenwarter, U., & Weidig, T. (2005). Exposed to the J-curve: Under- standing and managing private equity fund investments (discussion and bibliography). SSRN. https://doi.org/10.2139/ssrn.646845 Grilli, L., Mrkajic, B., & Lati, G. (2018). Venture capital in Europe: Social capital, formal institutions and mediation effects. Small Business Economics, 51(2), 393–410. https://doi.org/10.1007/ s11187-018-0007-7 Grinda, F. (2020, April 7). FJ Labs marketplace matrix. Fjlabs.Com. https://fabricegrinda.com/fj-labs-marketplace-matrix/ Grinda, F. (2021, February 19). FJ Labs valuation matrix. Fjlabs.Com. https://fabricegrinda.com/fj-labs-valuation-matrix/ Hallen, B. L., Cohen, S. L., & Bingham, C. B. (2020). Do accelerators work? If so, how? Organization Science, 31(2), 378–414. https:// doi.org/10.1287/orsc.2019.1304 Hayes, A. (2024, October 18). What is venture capital? Denition, pros, cons, and how it works. Investopedia. https://www.investopedia .com/terms/v/venturecapital.asp 128 ECONOMIC AND BUSINESS REVIEW 2025;27:115–129 Hayter, C. S., Nelson, A. J., Zayed, S., & O’Connor, A. C. (2018). Con- ceptualizing academic entrepreneurship ecosystems: A review, analysis and extension of the literature. The Journal of Technology Transfer, 43(4), 1039–1082. https://doi.org/10.1007/s10961-018 -9657-5 Hellmann, T., Montag, A., & Tåg, J. (2024). Tolerating losses for growth: J-curves in venture capital investing (IFN Working Paper No. 1500). Research Institute of Industrial Economics (IFN). https://hdl.handle.net/10419/302271 Isabelle, D. (2013, February). Key factors affecting a technology entrepreneur’s choice of incubator or accelerator. Technology In- novation Management Review, 16–22. https://doi.org/10.22215/ timreview/656 Kaplan, S. N., & Schoar, A. (2005). Private equity performance: Re- turns, persistence, and capital ows. The Journal of Finance, 60(4), 1791–1823. https://doi.org/10.1111/J.1540-6261.2005.00780.X Kim, J. H., & Wagman, L. (2014). Portfolio size and information dis- closure: An analysis of startup accelerators. Journal of Corporate Finance, 29, 520–534. https://doi.org/10.1016/j.jcorpn.2014.10 .017 Kohler, T. (2016). Corporate accelerators: Building bridges between corporations and startups. Business Horizons, 59(3), 347–357. https://doi.org/10.1016/j.bushor.2016.01.008 Kupp, M., Marval, M., & Borchers, P . (2017). Corporate accelerators: fostering innovation while bringing together startups and large rms. Journal of Business Strategy, 38(6), 47–53. https://doi.org/ 10.1108/JBS-12-2016-0145 Leitão, J., Pereira, D., & Gonçalves, Â. (2022). Business incubators, accelerators, and performance of technology-based ventures: A systematic literature review. Journal of Open Innovation: Technol- ogy, Market, and Complexity, 8(1), Article 46. https://doi.org/ 10.3390/JOITMC8010046 Meyer, T., & Mathonet, P .-Y. (2005). Beyond the J curve: Managing a portfolio of venture capital and private equity funds. John Wiley & Sons. Mian, S., Lamine, W., & Fayolle, A. (2016). Technology business incubation: An overview of the state of knowledge. Technova- tion, 50–51, 1–12. https://doi.org/10.1016/j.technovation.2016 .02.005 microcap.co. (2021, March 1). Startup valuation revenue multiple. https://microcap.co/startup-valuation-revenue-multiple/ Miller, C. (2022, March 18). SaaS valuation in 2024: How to value a SaaS company. Mosaic. https://www.mosaic.tech/post/saas -valuation Miller, P ., & Bound, K. (2011). The startup factories: The rise of accelerator programmes to support new technology ventures [Dis- cussion paper]. NESTA. https://www.nesta.org.uk/report/ the-startup-factories/ Moreira, H. (2023, May 25). TVPI vs. MOIC: Do you understand the difference? Financestu. https://nancestu.com/tvpi-vs-moic/ Moroz, P . W., Sierra, O., & Anderson, R. (2024). A structured review of start-up accelerator performance measurement: An integrated entrepreneurial program evaluation approach. En- trepreneurship & Regional Development, 36(3–4), 460–488. https:// doi.org/10.1080/08985626.2023.2298974 Ostin, O. (2021, January 22). A K-shaped VC market: Is the era of 10-20x revenue multiples here to stay? Crunchbase. https://news.crunchbase.com/startups/k-shaped-economic -recovery-startups-vc/ Pauwels, C., Clarysse, B., Wright, M., & Van Hove, J. (2016). Understanding a new generation incubation model: The ac- celerator. Technovation, 50–51, 13–24. https://doi.org/10.1016/ j.technovation.2015.09.003 Prencipe, D. (2017). The European venture capital landscape: An EIF perspective. Volume III: Liquidity events and returns of EIF-backed VC investments (EIF Working Paper). European Investment Fund (EIF). https://hdl.handle.net/10419/176671 Pustovrh, A. (2019). Sistem pospeševanja rasti inovativnih pod- jetij v zagonskih pospeševalnikih v manj razvitem podpornem okolju [System for accelerating the growth of innovative compa- nies in startup accelerators in less developed support environ- ments]. Economic and Business Review, 21(4), 173–179. https:// doi.org/10.15458/2335-4216.1075 Pustovrh, A., Rangus, K., & Drnovšek, M. (2020). The role of open innovation in developing an entrepreneurial support ecosys- tem. Technological Forecasting and Social Change, 152. https:// doi.org/10.1016/j.techfore.2019.119892 Radojevich-Kelley, N., & Hoffman, D. L. (2012). Analysis of accel- erator companies: An exploratory case study of their programs, processes, and early results. Small Business Institute Journal, 8(2), 54–70. Regmi, K., Ahmed, S. A., & Quinn, M. (2015). Data driven analysis of startup accelerators. Universal Journal of Industrial and Busi- ness Management, 3(2), 54–57. https://doi.org/10.13189/UJIBM .2015.030203 Sahlman, W. A. (1990). The structure and governance of venture- capital organizations. Journal of Financial Economics, 27(2), 473– 521. https://doi.org/10.1016/0304-405X(90)90065-8 Sarath, C. P . (2021, October 21). Why your startup should care about revenue multiples HackerNoon. https://hackernoon.com/why -your-startup-should-care-about-revenue-multiples Smith, S. W., Hannigan, T. J., & Gasiorowski, L. (2016). Peering inside: How do peer effects impact entrepreneurial outcomes in acceler- ators? https://mackinstitute.wharton.upenn.edu/wp-content/ uploads/2016/03/Winston-Smith-Sheryl-Hannigan-Thomas -and-Gasiorowski-Laura_Peering-Inside.How-do-Peer-Effects -Impact-Entrepreneurial-Outcomes-in-Accelerators.pdf Stayton, J., & Mangematin, V . (2019). Seed accelerators and the speed of new venture creation. The Journal of Technology Transfer, 44(4), 1163–1187. https://doi.org/10.1007/s10961-017-9646-0 Swildens, H., & Yee, E. (2017, February 7). The venture capital risk and return matrix. Industry Ventures. https:// www.industryventures.com/insight/the-venture-capital-risk -and-return-matrix/ Tekic, Z., Hrynkevich, A., & Malyy, M. (2024). Do accelerators promote the growth of startups? Analysing the effectiveness of startup accelerators through the lens of big data. Technol- ogy Analysis & Strategic Management. https://doi.org/10.1080/ 09537325.2024.2383608 Uhm, C. H., Sung, C. S., & Park, J. Y. (2018). Understanding the accelerator from resources-based perspective. Asia Pacic Jour- nal of Innovation and Entrepreneurship, 12(3), 258–278. https:// doi.org/10.1108/APJIE-01-2018-0001 Weiblen, T., & Chesbrough, H. W. (2015). Engaging with startups to enhance corporate innovation. California Management Review, 57(2), 66–90. https://doi.org/10.1525/cmr.2015.57.2.66 Weidig, T., & Mathonet, P .-Y. (2004). The risk prole of private equity . SSRN. https://doi.org/10.2139/SSRN.495482 Wilhelm, A. (2021, July 14). How to make the math work for today’s sky-high startup valuations. TechCrunch. https:// techcrunch.com/2021/07/14/how-to-make-the-math-work -for-todays-sky-high-startup-valuations/ Wise, S., & Valliere, D. (2014, Winter). The impact on management experience on the performance of start-ups within accelerators. The Journal of Private Equity, 18, 9–19. https://www.jstor.org/ stable/43503826