Innovative financing channels: International evidence from initial coin offering and project start-up success factors

Ftiti, Zied; Hamza, Taher; Mnif, Zouhour; Louhichi, Waël

doi:https://doi.org/10.59876/a-3q84-6zn7

Firms are substantially faced with innovations of various natures, subjecting them to risk and uncertainty, and leading them to deal with a complex environment. Funding channels are one of the main challenges in the face of global competition and technological change, leading firms to seek external partners to develop different forms of inter-organizational alliances. Initial coin offerings (ICOs) are drawing increasing interest as an alternative to conventional funding channels, such as angel investors or venture capitalists. They represent a novel type of venture financing through cryptocurrencies and blockchain technology (Catalini & Gans, 2018; Roosenboom et al., 2020). In the early stages of their business growth, blockchain-based companies raise funds from individual investors via information and communication technologies (e.g., websites or social media). In the context of the relationship between the ICO project team and ICO investors, the latter appear to exhibit characteristics similar to those of innovation adopters and contribute to the development of financing innovation. ICOs are considered to be a recent innovation that leads a firm to have an alliance with other partners, in order to allow ICOs to be accepted and tradeable. Factual observations show that the digital token selling mechanism based on blockchain technology is a breakthrough in start-up innovation financing that has spread globally. Given the high costs of research and development, this type of funding can be significant for highly innovative businesses.

Benedetti and Kostovetsky (2021) showed that in the 16 months since January 2017, ICOs were able to raise more than $12 billion. For each ICO, entrepreneurs release a whitepaper that contains information related to their ICO to provide information and attract potential contributors. During ICOs, entrepreneurs seek to raise funds for their own ventures by creating tokens[1] and selling them to investors in exchange for cryptocurrencies (e.g., Bitcoin and Ethereum) or fiat currency (e.g., US Dollars). The collected funds are used to finance projects that are still in the idea stage. After completion of the ICO, tokens may be traded on an online exchange platform, and investors have the choice to sell or hold on to these tokens. There is no rule defining how long an ICO should last. Some ICOs last less than one day, whereas others continue for more than a year. Apart from the fact that ICOs were banned in South Korea and China toward the end of 2017, ICO regulations around the world are not comprehensively defined. Each country is free to decide whether to regulate ICOs or not. However, ICOs present some risks (e.g., scams, hacking) that are considered a real danger for investors and entrepreneurs.

Nevertheless, ICOs are showing rapid growth on a global scale and help raise funds in support of small- and medium-sized enterprises (SMEs) and start-ups. Hence, this innovative financing channel deserves research focus. In particular, several issues related to ICOs require investigation as they have implications on the decisions of market investors and the process of raising alternative funds serving SMEs and start-up managers. A strand of literature is dedicated to elucidating the distinct nature of ICOs as compared to other entrepreneurial financing channels like crowdfunding. Lipusch (2018) delved into the definition of ICOs and contrasted them with other traditional fundraising channels, including crowdfunding, venture capital, and Initial Public Offerings (IPOs). Complementing this work, Block et al. (2021) elaborated on the nuanced differences between ICOs and crowdfunding, focusing on aspects such as stakeholder participation, platform utilization, market operations, and regulatory oversight. As several companies have successfully raised funds through ICOs, with some of their tokens being traded on online exchange platforms, the literature has evolved to investigate the factors driving the success of ICOs, focusing on different perspectives. Varnaitė (2018) examined the factors contributing to the success of ICOs, from investors’ perspective. Through interviews with ICO experts, the study revealed that investors are solely driven by the objective of achieving a profitable financial return on their investment. Their results showed that ICO experts ascribe a greater probability of success to the qualities of the management team and specific team attributes, rather than to the inherent merits of the business itself. For example, Adhami et al. (2018) investigated the determinants of ICO success and identified them to be code source availability, the organization of the token presale, and the offering of specific services for contributors. Burns and Moro (2018) reported that team quality influences the success of ICOs. Fisch (2019) found that a high-quality source code stimulates the increase in the amounts of funds raised through ICOs. Howell et al. (2020) highlighted credible project commitment, quality signals, and disclosure as the main success factors of ICOs. Belitski and Boreiko (2022) showed that the presale of tokens, bonus sales, and funders’ ownership contribute to ICO performance. Meanwhile, Gächter and Gächter (2021) identified lucky timing as a crucial success factor.

As ICO might be successful only in the short term, the literature has evolved to assess the factors driving long-run post-ICO performance. Roosenboom et al. (2020), based on the 630 ICOs ranging from 2015 to 2017, showed that disclosure quality to investors, higher quality rating, bonus or a discount to early investors, and size of venture team are the factors that matter. Ahmad et al. (2023), based on 503 ICOs ranged between 2015 and 2018, showed that the main factors driving the success of the post-ICO performance are insider retention, resource-related signals, such as the number of team members and advisors, bonus or discount to early investors, presale offers, and country’s financial system development. Aslan et al. (2023) investigated the ICO success factors and aftermarket performance of ICOs. In addition to higher ratings, shorter planned token sale duration, smaller share for token sale, and larger number of experts, the authors showed that market sentiment significantly explains the post-ICO performance. Sapkota and Grobys (2023) found that ICO investors are guided by emotions when making investment decisions rather than the quality and readability of ICO white papers. Other studies have investigated the success factors explaining post-ICO performance. The literature dealing with ICO success factors and those explaining the post-ICO performance are scarce and inconclusive. These studies employ various factors and look at the different perspectives of ICO success factors from investors perspective (Adhami et al., 2018; Varnaitė, 2018), agility perspective (Lipusch, 2018), governance perspective (Giudici & Adhami, 2019), team characteristics (Ahmad et al., 2023; Amsden & Schweizer, 2018; Roosenboom et al., 2020), and voluntary disclosure quality (Chitsazan et al., 2022; Fisch, 2019; Lyandres et al., 2022; Momtaz, 2021; Sapkota & Grobys, 2023).

Our study aims to extend this recent and scarce literature by investigating both the factors driving ICO success as well as the factors explaining post-ICO performance. More practically, as previous literature used various factors, this study attempts to disaggregate factors across four strands: voluntary disclosure quality, team characteristics, deal characteristics, and country of issuance or geographic location of ICO. In our analysis, we consider two types of tokens—security and payment. Our sample comprises 410 completed, traded, and nontraded ICOs that have different types of tokens and are from different countries in the 2016–2018 period. Our data are handpicked from different databases, especially the ICObench website. As our dependent variables are dummy and quantitative variables, we apply ordinary least squares (OLS) and logistic regressions depending on the type of variable used in the analysis. Our study contributes to the emerging literature on ICOs in at least three ways. First, it proposes a comprehensive typology of potential determinants of ICO’s success. Second, our study provides new empirical results about the success of project start-ups. From the perspective of reducing asymmetric information and investor adverse selection, our study shows the significant effect of expert and transparency ratings on ICO success and project start-up success. All the other voluntary disclosure quality variables point in the same direction. Third, this study shows the importance of project team, in particular, the negative perception of the gender factor. Investors do not seem to appreciate women’s low-risk profile when launching and running a start-up.

The remainder of this article proceeds as follows. Section 2 reviews the literature on the determinants of ICO success. Section 3 describes the sample and methodology. Section 4 reports the summary statistics, empirical results, and robustness checks. Section 5 concludes the paper.

Literature review

In the following section, we offer an analytical review of empirical findings organized according to the categorization of success factors employed in our research: voluntary disclosure quality, team characteristics, deal characteristics (both supply and demand aspects), and country of issuance. ICOs are characterized by a high level of information asymmetry because ventures issuing ICOs are generally in the early stages. The lack of formal disclosure mandates, coupled with a heightened preference for anonymity, exacerbates this issue by limiting access to reliable information, thereby escalating the risk of fraudulent activities (Shifflett & Jones, 2018). To elucidate the driving factors behind ICO success, we draw upon signaling theory (Spence, 1973). As emphasized by Chitsazan et al. (2022) in their systematic analysis of 78 empirical studies, most studies on ICO success used signaling theory to examine its determinants. This theoretical setting provides an appropriate framework for determining how firms can minimize information asymmetry between the key players (founders and investors) in the ICO process. Elevated levels of information asymmetry typically necessitate stronger signaling mechanisms. Within this context, the principles of voluntary disclosure and signaling theory serve as effective frameworks for understanding how firms can mitigate information asymmetry by strategically deploying voluntary disclosures and other signaling methods to attract increased capital (e.g., Bourveau et al., 2022; Chitsazan et al., 2022; Momtaz, 2021; Chod & Lyandres, 2021; Howell et al., 2020; Giudici & Adhami, 2019; Yermack, 2015). Resource-based theory complements this theoretical framework as it provides a lens through which various success factors of ICOs and project start-ups post-ICO can be analyzed. It considers how the unique resources, capabilities, and competencies of these ventures contribute to their competitive advantage, sustainability, and ability in a rapidly changing environment (Fisch & Momtaz, 2020; Hackober & Bock, 2021; Schaefer & Strese, 2021; Xu et al., 2021). This perspective offers a unifying framework to understand how projects navigate the complexities of both the ICO phase and the post-ICO operational phase.

Voluntary disclosure quality

The literature on the voluntary disclosure of fundraising activities stresses that firms are inclined to provide information voluntarily to stakeholders and investors to reduce information asymmetry regarding the quality of the project supported by the management team (adverse selection), which reduces risk-taking by these investors and increases the probability of ICO success. Roosenboom et al. (2020) explored the determinants that influence ex-post-ICO success. Their results indicate that ICOs with higher expertise and transparency ratings[2] significantly influence success post-ICO. The authors showed that having deposits on GitHub prior to the ICO significantly and positively affects ex-post-ICO success. Cohney et al. (2018) argued that having a code published online on GitHub, an open-source community platform for programmers that allows investors to monitor the progress of codes, could signify technological capabilities that positively impact ICO success. Furthermore, Cohney et al. (2018) considered that social media presence facilitates access to information required by investors and reduces information asymmetry. Belitski and Boreiko (2022), Fisch (2019), and Howell et al. (2020) showed a positive and highly significant association between the issuance of a whitepaper (e.g., information about a project’s business plan, blockchain platform, founding team profile), in which information is disclosed voluntarily, and ICO success. Fisch (2019) and Amsden and Schweizer (2018) documented the positive effect of the length of the whitepaper on ICO success. Features such as the Know Your Customer[3] (KYC) procedure and the process of whitelisting[4] have been implemented to secure the ICO market from fraudulent activities. Chen (2019) examined the nature and impact of signals disseminated through various channels at different stages of an ICO to explore the relationship between signal quality and information asymmetry. The study finds that during the crowd sale phase, highly credible and easily interpretable signals exert a significant influence on token sales. Conversely, during the listing stage, signals that are of lower credibility yet easily interpretable have a marked impact on token trading. Notably, highly credible but difficult-to-interpret signals, which often convey fundamental project information, lose their efficacy in both stages, thereby contributing to information asymmetry within the ICO environment. Bourveau et al. (2022) explored the influence of disclosure practices and information intermediaries within the unregulated crypto-token market, focusing their analysis on a sample of 2,113 ICOs. Their findings indicate that firms with more comprehensive disclosure practices are more likely to successfully raise funds. Intriguingly, this relationship is amplified when credibility-enhancing mechanisms are in place, such as robust internal governance structures or external validation from information intermediaries.

Team characteristics

Another corpus of studies is interested in team quality, such as team size, years of experience, and the gender of members. Sapkota and Grobys (2023), Lyandres et al. (2022), Ahmad et al. (2023), Roosenboom et al. (2020), Amsden and Schweizer (2018) showed that a larger team size and token retention are positively related to ICO success. Johnson et al. (2018) found that the presence of a female entrepreneur is positively related to the amount of funds raised. Howell et al. (2020) provided evidence that “the issuer entrepreneurial experience” has a positive and significant relationship with ICO success. Ahmad et al. (2023), while investigating the factors of post-ICO success, found that a large number of advisors and staff members is significantly and positively associated with the probability of successful ICO post-performance. Lyandres et al. (2022) investigated the case of 7,514 unique merged projects in 133 countries for both ICO success and post-ICO success. They showed that size of venture team is among the main factors driving the ICO success. However, Sapkota and Grobys (2023) showed that team size marginally impacts the ICO success.

Deal characteristics

Many scholars have investigated the determinants of ICO success. Some of these works have indicated that the major determinant of success is the completion of a pre-ICO phase (ICO presale). Presales often serve to cover upfront costs, such as the cost of paying for promotional ads. However, Ahmad et al. (2023), and Amsden and Schweizer (2018), found a negative relationship between presale activities and ICO success and suggested that a presale may indicate that the venture team holds insufficient funds to cover their costs. By contrast, Roosenboom et al. (2020) and Adhami et al. (2018) showed that the probability of ICO success is higher if a presale is conducted. Meanwhile, Lyandres et al. (2022) found that a presale has an insignificant effect on ICO success.

Another factor that could potentially influence the success of ICOs is the implementation of a bonus scheme, where discounts are offered to investors during the ICO phase. In the context of crowdfunding, Lyandres et al. (2022) and Amsden and Schweizer (2018) argue, in line with Hornuf and Schwienbacher (2018), that such bonus schemes have a significantly positive relationship with token tradability. These studies suggest that bonus schemes may encourage early investment, thereby increasing the liquidity of the token once it is tradable. Meanwhile, Ahmad et al. (2023) and Roosenboom et al. (2020) presented a contrasting view, asserting that the absence of a bonus structure during the ICO phase is positively and significantly correlated with ICO success. These studies argued that not offering bonuses may signal to investors that the ICO is confident in its value proposition and does not need to rely on incentives to attract capital. The absence of a bonus scheme may also reduce the risk of “pump-and-dump” schemes, where investors take advantage of bonuses by selling immediately, once the token becomes tradable, thereby undermining its long-term value.

The existent ICO literature also considers the price of tokens and the total token supply as factors influencing ICO success. Benedetti and Kostovetsky (2021), Burns and Moro (2018), and Roosenboom et al. (2020) showed a negative association between token price and ICO success, which suggests that lower token prices during ICOs serve as a good signal for investors. In terms of token supply, Fisch (2019) provided evidence that unexpectedly, ICOs issuing a higher number of tokens lead to ICO success. A possible explanation is that a high supply may have similar characteristics to a lottery and thus attract more investors and funds.

Regarding the number of cryptocurrencies, Amsden and Schweizer (2018) emphasized that the acceptance of fiat currencies for payment is considered a determinant of success. They argued that ICOs accepting several currencies need significant expertise to incorporate functionalities between them. Accepting fiat currencies for payment positively and negatively affects success. The positive association with the funds raised is explained by the fact that the acceptance of fiat currencies shows the link between ICO funding channels and the conventional banking system. Amsden and Schweizer (2018) supported the positive effect of using fiat currencies on ICO success. A possible explanation for the negative impact on token tradability is that fiat currencies can carry a cost for the venture. The authors argued that more accepted currencies increase the chances for a higher score from ICObench because of their variety and availability for investors. The duration of the ICO is also considered a principal determinant of ICO success. It represents the length of the ICO phase. Fisch (2019) and Roosenboom et al. (2020) stated that planning a shorter duration for the ICO positively and significantly impacts ICO success, as reported in the literature on crowdfunding (Courtney et al., 2017; Mollick, 2014).

Roosenboom et al. (2020) analyzed target amounts (soft cap and hard cap) as an explanatory variable as they may influence ICO success and post-ICO performance. They reported that setting greater target amounts negatively impacts ICO success. Furthermore, Fisch (2019), Howell et al. (2020), and Amsden and Schweizer (2018) showed that there is a significant and positive correlation between tokens based on the Ethereum platform and ICO success. However, Burns and Moro (2018) documented that the use of the Ethereum platform is significantly and negatively associated with success as the largest ICOs have to establish their own platforms.

Fisch (2019) reported that intense Twitter activity during the ICO is related to a higher amount of funding raised, thereby serving as a measure of ICO success. In the same vein, Howell et al. (2020), Benedetti and Kostovetsky (2021), and Burns et Moro (2018) found a positive correlation between the number of Twitter followers and ICO success. Yadav (2017) explained that the process of KYC is a crucial indicator of success in the ICO context. In line with this, Lyandres et al. (2022) identified a positive effect of using the KYC process on ICO success.

Country of issuance or geographic location of ICO

Few empirical studies have investigated the drivers of ICO activities across countries and assessed their impact on ICO success. For entrepreneurs and project owners, understanding ICOs’ geographic locations helps determine where to initiate an ICO as it leads to the identification of the most appropriate regulatory, fiscal, economic, and financial environments to raise funds. According to the existing literature, innovative activities evolve in tandem with capital markets and financial systems, as well as the technological and legal environments. Ahmad et al. (2023) found a strong and positive relationship between legal friendliness and ICO activity. Similarly, Huang et al. (2020) argued that countries with ICO-friendly regulations have more ICOs and that tax regimes are not clearly related to ICOs. They also emphasized that ICOs take place more frequently in countries with developed financial systems, public equity markets, and advanced digital technologies.

In Table 1, we present a summary of the main literature on the determinants of successful ICOs based on the proposed typology.

Table 1

**Review of recent literature**

Data and methodology

Sample selection and data

Our initial sample comprised 410 completed ICOs listed predominantly on the ICObench website and other websites over the 2016–2018 period. Figure 1 shows the distribution of ICOs across time.

Figure 1

**Distribution of ICOs across time**

As there is no single database that contains all information relative to ICOs, we handpicked our ICO data across 12 different databases[5]. Data were gathered initially from ICObench and ICOmarks and then cross-checked against those in other databases. ICObench is the database that provides most of the high-quality data and information on ICOs and may thus be considered as the “core of our sample.” Although databases are considered comprehensive (taken together), they still lack key information that we compiled manually from ICOs’ websites and whitepapers. After the completion of a public sale, many ICOs remove information from databases and even disable their websites and social media accounts (e.g., Twitter). This can be a sign of failure. As we could find sufficient information on some ICOs, we were forced to omit them.

Our sample comprised listed ICOs on an online exchange platform for trading and non-listed ICOs. We used the website CoinMarketCap (CMC) to distinguish tradable ICOs from nontradable ones. Figure 2 hereafter shows the geographical distribution of our selected ICOs. The USA has the highest number of ICOs (74 of 410 ICOs) and thus represents 18% of our sample. It is followed by Singapore with the second-highest percentage of 10.5% of the ICOs from our sample (43 of 410 ICOs). Russia comes in third with 8% of the ICOs (33 of 410 ICOs). The other ICOs are in the UK (7.8%), Switzerland (7.07%), Estonia (4.63%), and so on. The remaining ICOs (31.5%) are distributed over several countries worldwide with very low rates.

Methodology

Dependent variables

Our study investigates the determinants of ICO and project start-up success. For ICO success, we considered the following four measures of ICO success as dependent variables, based on prior studies:

Soft cap hit: A dummy variable indicating whether the minimum goal (soft cap) is reached,

Funding percentage: The percentage of the raised amount from the maximum goal (hard cap),
Ln (amount initially raised during the ICO): The natural logarithm of the raised amount,
Token tradability: A dummy variable indicating whether the ICO is listed in an online exchange platform tracked by CMC[6].
For post-ICO success, we considered the following four measures as dependent variables:

Website online status: Indicates whether the ICO website is online in the six months following the end of the ICO,

Tweets per week: The number of tweets per week over the period between the start of the ICO and six months later,
Twitter activity: A dummy variable that indicates whether the ICO’s Twitter account is active over the period between the start of the ICO and six months later,
GitHub activity: A dummy variable indicating whether there are contributions to repositories on GitHub in the last period.

Explanatory variables

According to the previous literature, this study proposes a comprehensive analysis of all potential determinants of ICO success and a typology of these factors: i) voluntary disclosure quality, ii) team characteristics, iii) deal characteristics (offer and demand characteristics), and iv) country of issuance or geographic location of ICO.

Table 2 presents all variable definitions and the expected sign of the different explanatory variables.

Table 2

**Variable definitions and expected signs**

Econometric models

Based on the correlation matrix (unreported) and variance inflation factor (VIF), we considered two sets of models to avoid issues related to multicollinearity. The first set of models is dedicated to expert and transparency ratings as independent variables that are highly related to other variables. We chose to test their impact on ICO and project start-up success in a univariate framework, presented in models (1) and (2), respectively.

Our second set of models is dedicated to the explanatory variables in a multivariate analysis framework. We estimated the following model to examine the determinants of ICO and project start-up success, as presented in models (3) and (4) respectively.

We used OLS regressions for the quantitative dependent variables and logistic regressions for the dummy dependent variables.

Summary statistics and empirical results

Summary statistics

Table 3 reports the summary statistics. It shows that the average raised amount is $21.9 million with a minimum amount of 0 and a maximum amount of $2,248.750 million related to an ICO entitled “ClearPoll” in October 2017. In our sample, 73.66% of the ICOs had tradable tokens. Table 1 also indicates that 259 ICOs, representing 63% of our sample, fixed their soft caps. The average soft cap equals $8.75 million with a maximum of $346 million. The average hard cap is $48.5 million, with the maximum and minimum being $3.21 billion and $300,000, respectively. Of the 410 ICOs, 365 (89%) had online websites in February 2019. The average number of tweets per week in the period between the start of the ICO and February 2019 was 11.76 tweets. Out of our sample of ICOs, 81.22% or 333 ICOs were active on Twitter while only 32.2% or 132 ICOs were active on GitHub from November 2018 to February 2019.

Regarding the independent variables, Table 3 shows that the average expert rating is 3.53 and that the median equals 3.6. The average transparency ratings equal 3.44, and the median is 3.7. Half of the ICOs in our sample had repositories on GitHub before launching their public sale. The teams retained 13.77% of the issued tokens on average, with the maximum and minimum being 50% and 0%, respectively. The percentage of sold tokens was 56.15% of the issued tokens on average, with the maximum and minimum being 100% and 3%, respectively. Presales were noted in 67.32% of our sample. Bonus schemes were set up in 251 cases (61.22% of the sample). Table 3 shows that the planned duration of ICOs is almost 32 days with a median of 30 days. Team size ranged from 2 to 132 members, with the average being 11 members. Finally, female team members were noted in 69.15% of our sample.

Empirical results

Univariate analysis: The effect of expert and transparency ratings

We investigated the effect of expert and transparency ratings on ICO and project start-up success. We conducted OLS and the logit regressions by using expert and transparency ratings separately as independent variables.[7] Tables 4 and 5 report the results of our models for ICO success and project start-up success, respectively.

The results showed that the expert rating (Panel A, Table 4) and transparency rating (Panel B, Table 4) are not significant determinants of ICO success, except when using the funding raised as a proxy for ICO success with coefficient equal to 0.3265 significant at 5%. A possible explanation for our results is that the ratings change every few hours (during the ICO). We also note that our selected ICOs are completed ICOs. Thus, we conclude that these ratings are appropriate for the post-ICO period. This is shown in the results provided in Table 5, which reports the effect of expert rating and transparency rating on project start-up success. Thus, both types of ratings exert a significant and positive effect on project start-up success, panels A and B of table 5 respectively. This finding is in line with those of Aslan et al. (2023) and Roosenboom et al. (2020), who reported a positive and significant effect of both ratings on success.

Table 4

**OLS and logit regressions for ICO success: Expert and transparency ratings**

Table 5

**OLS and logit regressions for project start-up success: Expert and transparency ratings**

Multivariate analysis: Determinants of success

We examined the determinants of ICO and project start-up success. In this study, the regressions did not suffer from any multicollinearity issues as the VIF coefficients are lower than 3. Table 6 provides the results of our regressions using the four measures for ICO success as dependent variables: soft cap hit, funding percentage, funding raised, and token tradability.

Our empirical results showed that overall, ICO success is explained by the following factors: voluntary disclosure quality, team characteristics, deal characteristics (offer and demand characteristics), and country of issuance or geographic location of ICO.

Concerning voluntary disclosure quality, we found that the presence of repositories on GitHub has a positive effect on the amount of funds raised and token tradability. This result is in line with those of Cohney et al. (2018). It indicates that posting the code on GitHub before the beginning of the ICO signals a high level of confidence in the project. Moreover, the number of whitepaper pages positively and significantly impacts the soft cap hit and the amount of funds raised. However, it is important to discuss that GitHub disclosure might be a source of risk, leading the companies to adopt some specific operational approaches as risk mitigation strategies. For example, they might use private repositories for sensitive data and code, limited access permissions, and secure encryption methods. Furthermore, companies may choose to only disclose certain parts of their codebase that are less sensitive but still indicative of their technical prowess, thus maintaining a level of information asymmetry that protects their competitive edge. The presence of a whitepaper negatively impacts the funding raised. A possible explanation for this result is that we think that some ICOs deleted their whitepapers once the ICO ended. Hence, we did not have information about whether the ICOs published whitepapers and then deleted them or never posted one. Therefore, given that the ICOs in our sample are completed ICOs, we had to consider that the nonexistence of an online whitepaper indicated that one was never published.

For team characteristics, we found that a larger team positively influences the amount of funds raised. Such finding is similar to those of Ahmad et al. (2023), Amsden and Schweizer (2018), and Roosenboom et al. (2020). Interestingly, we observed that having a female team member negatively and significantly impacts the ICO success measured by different proxies: the probability of reaching the soft cap amount, the amount of funds raised, and the percentage of funds raised in hard cap. Our result contrasts with those of Johnson et al. (2018) who found the opposite relationship. According to behavioral finance, our result is explained by the excess of confidence among male investors. The more the domain is considered “masculine,” the more men feel touched by divine grace. As 90% of crypto investors are males (Charness & Gneezy, 2012), those who invest in ICOs tend to prefer ICOs with more male team members. Therefore, this negative association between female members and ICO success reflects a negative perception by investors of the gender factor, which can be attributed to the fact that investors do not seem to appreciate the low-risk profile of women when launching and managing a start-up. Finally, we found that leadership has a significant and positive impact on the funding raised. This shows the importance of a positive perception by investors of the leadership of the project team.

Table 6

**OLS and logit regressions for ICO success: Determinants of ICO success**

Our results concerning deal characteristics showed that setting a large soft cap amount reduces the likelihood of hitting it and setting a large hard cap amount decreases the percentage of funds from this amount. In line with our expectations, a shorter ICO duration significantly influences ICO success. In fact, a shorter duration signals the capability of attaining the amount of funds needed quickly. Our findings are aligned with those of Fisch (2019) and Roosenboom et al. (2020). We also found that accepting several cryptocurrencies for payment positively and significantly affects the probability of reaching the soft cap amount, the amount of funds raised, and the tokens raised (Amsden & Schweizer, 2018). Meanwhile, accepting a fiat currency negatively influences the probability of reaching the soft cap amount. This result supports the idea that crypto investors are unable to complete ICOs on their own. It is in line with that of Amsden and Schweizer (2018), who found that not accepting fiat currency for payment and having a complete whitepaper positively influence success. Moreover, not setting up a bonus scheme significantly impacts the likelihood of reaching the soft cap amount. Such a situation signals the lack of confidence of entrepreneurs in their projects and their willingness to attract investors through the bonus structure. Moreover, issuing tokens on the Ethereum platform, having a long whitepaper, and being followed by a large number of users on Twitter exert significant and positive impacts on ICO success. Finally, it seems that setting a KYC process does not significantly impact the success of the ICO.

Regarding geographic location effect, our results showed that issuing tokens in Russia has a negative and significant association with the funding raised. Russia does not have friendly ICO regulations, and it has begun to set laws surrounding ICOs (Ahmad et al., 2023; Huang et al., 2020).

Table 7 provides the results from our multivariate regressions using the four measures for success after the completion of the ICO (project start-up period) as dependent variables (website online, tweets per week, Twitter activity, and GitHub activity).

Table 7

**OLS and logit regressions for project start-up success post-ICO**

Our empirical results showed that overall, project start-up success is explained by the following factors: voluntary disclosure quality, team characteristics, deal characteristics (offer and demand characteristics), and country of issuance or geographic location of ICO.

Our results showed that having repositories on GitHub pre-ICO positively and significantly impacted the GitHub activity in the recent period and the setting up of an online website post-ICO (Cohney et al., 2018). The results also showed that issuing a whitepaper is negatively associated with tweets per week.

Regarding the team characteristics, the results indicated that a higher percentage of tokens retained by insiders increases the probability of having an online website post-ICO. This relation highlights the high confidence of the insiders in the project. This result is similar to that of Roosenboom et al. (2020) and confirms the incentive and compensation theories. Moreover, the presence of a female team member has a significant negative impact only when the post-ICO success is measured by the probability of being active on GitHub. The gender variable is positive but not significant when we consider the other proxies of success after the completion of the ICO as dependent variables, that is, website online, tweets per week, and witter activity. This result shows that being a female negatively impacts the success of the ICO but does not have an impact on the post-ICO performance. This result can be explained by the excess of confidence and risk taking among male investors, which is a necessary condition for the success of ICOs. However, there is no gender effect on the managerial performance during the project start-up post-ICO.

The most significant result concerning deal characteristics showed a positive association between the number of followers on Twitter and ex-post ICO success. We also showed that accepting a high number of cryptocurrencies for payment, not allowing payments with fiat currency, and issuing a high number of tokens (supporting the risk sharing phenomenon) positively and significantly impact the project start-up success. We also confirmed that setting a KYC process does not impact the success during the pre- and post-ICO periods, except for a positive performance (evaluated by the presence of website) observed for the start-up period. Moreover, token price has a significant and negative association with Twitter activity. Finally, and as predicted, a shorter ICO duration impacts the GitHub activity in the recent period. Furthermore, presale seems to have a significantly negative association with tweets per week, but the overall effect on project start-up success is minimal.

Finally, after controlling for the country of issuance effect, we found that issuing tokens negatively impacts the GitHub activity in Russia and the USA and positively impacts that in Switzerland. Specifically, ICO issuance in Switzerland exhibits a positive and significant relationship with Twitter activity; this result was expected as Switzerland has friendly ICO regulations.

Robustness checks

To check the robustness of our results, we used generalized linear model (GLM) regressions for the quantitative dependent variables and probit regressions for the dummy dependent variables. In terms of the univariate analysis, the results of the probit and GLM regressions are presented in Tables 8 and 9 for ICO and project start-up success, respectively.

Table 8

**Robustness check: GLM and probit regressions for ICO success: Expert and transparency ratings**

As for the ICO and project start-up success based on expert and transparency ratings, the results are in line with those reported in Tables 4 and 5. They show a significant and positive association between expert rating and the funding raised as a proxy for ICO success. Meanwhile, the relation between transparency rating and ICO success does not show any significant difference. The expert and transparency ratings appear to be significantly and positively associated with project start-up success after the completion of the ICO. These results are in line with those reported in Tables 4 and 5.

Table 9

**Robustness check: GLM and probit regressions for post-ICO success: Expert and transparency ratings**

In terms of the multivariate analysis, the results of the probit and GLM regressions are presented in Tables 10 and 11 for ICO and project start-up success, respectively. The results do not deviate significantly from those reported in Tables 6 and 7, although some variables become more significant. Furthermore, the results globally confirm those reported in Tables 5 and 6 for ICO and project start-up success, respectively.

Table 10

**Robustness check: GLM and probit regressions for ICO success**

Table 11

**Robustness check: GLM and probit regressions for project start‑up success post-ICO**

Conclusion

Our study is among the first to examine the determinants of ICO and project start-up success. The ICO is a novel financing channel that enables start-ups to raise funds for their venture projects by selling newly created tokens. We investigate the factors that contribute to the success of ICOs and project start-ups post-ICO.

Using a sample of 410 ICOs over the 2016–2018 period, we empirically showed that the following factors exhibit a positive and significant effect on ICO success: i) disclosure quality (GitHub, whitepaper pages); ii) project team characteristics (team size, leadership, gender, diversity); and iii) certain deal characteristics, such as subscription duration and the number of accepted cryptocurrencies. Additionally, the success of project start-ups post-ICO is affected by insider ownership, expert and transparency ratings, and country of token issuance. Our empirical results are robust to the use of alternative regression methods and offer several practical implications.

Our study contributes to the emerging literature on ICOs in at least three ways. First, it proposes a comprehensive typology of potential determinants of ICO success. Second, it provides new empirical results about the success of project start-ups. From the perspective of reducing asymmetric information and investor adverse selection, our study shows the significant effect of expert and transparency ratings on ICO success and project start-up success. All the other voluntary disclosure quality variables point in the same direction. Third, this study shows the importance of the project team, in particular, the negative perception of the gender factor. Investors do not seem to appreciate women’s low-risk profile when launching and running a start-up.

In addition, this study offers several noteworthy managerial implications. For example, it allows a better understanding of the links between the characteristics of ICOs and their success during and after deal transactions. This could encourage start-ups to use this new channel of fundraising to finance their projects in the early stages. Thus, entrepreneurs can identify key success factors during and post-ICO to attract investors. In addition, investors will be able to make arbitrage portfolios, targeting ICOs that have the highest probability of success and are potentially value-creating.

Our research findings can be further contextualized within the broader scope of entrepreneurial finance literature, particularly the studies focusing on fundraising channels. Notably, our results indicate that factors such as diversity within the project team and the country of issuance play a significant role in both the immediate success and project start-up post-ICO. This opens up intriguing possibilities for examining various country-specific characteristics, such as the governance structures and decision-making processes of ICO-funded projects and their contribution to investor confidence. How do voting mechanisms and governance token dynamics influence project direction and success? Based on our empirical findings, future research could conduct comparative studies across different ICOs and blockchain ecosystems to identify common patterns and success factors, and explore regional variations in ICO success and regulatory approaches. Finally, another promising direction for future research would be examining the growing importance of environmental and sustainability factors in ICOs, including the adoption of eco-friendly consensus mechanisms and the carbon footprint of blockchain projects.

Abstract

Résumé

Resumen