quality revealing or overstating? -...
TRANSCRIPT
Quality Revealing or Overstating?
—Analysis on Qualitative Startup Information in Equity Crowdfunding
Sofia Johan
Tilburg Law and Economics Centre (TILEC)
and
York University - Schulich School of Business
4700 Keele Street
Toronto, Ontario M3J 1P3
Canada
Yelin Zhang
School of Business Administration - Gonzaga University
502 East Boone Avenue
Spokane, WA 99258-0102
* We are indebted to the mutual funds for providing access to their detailed data. Also, we are indebted to the
Canadian Securities Administrators and the Ontario Securities Commission, and in particular Minami Ganaha,
Rhonda Goldberg, Chantal Mainville, Paul Redman, and Dennis Yanchus, for their help with facilitating data
collection. We also owe thanks to the seminar participants at Exchange Traded Funds Conference (Ottawa),
Fidelity Financial, FMA Europe (Helsinki), Hong Kong Polytechnic University, Investor’s Group, Northern Finance
Association, Ontario Securities Commission, the University of Birmingham, the University of Exeter, and York
University. We owe thanks to the Ontario Securities Commission and the Social Sciences and Humanities Research
Council of Canada for financial support.
Quality Revealing or Overstating?
—Analysis on Qualitative Firm Information in Equity Crowdfunding
Abstract
This paper studies the impact of qualitative business information on mitigating
information asymmetry between equity crowdfunding entrepreneurs and investors. Qualitative
business information covers entrepreneurs’ introduction on business model, competitive strategy,
product market, drivers and barriers for product/service adoption and business milestones.
Empirical data reveal that more detailed disclosure of qualitative business information leads to
better fundraising outcome. It pays off for all startups to disclose qualitative information,
although high quality firms benefit more from the disclosure than low quality firms. In addition,
entrepreneurs’ excessive use of promotional language, or self-praise on quality without factual
support, is not rewarded by investors.
Key Words: Equity Crowdfunding, Information Asymmetry, Qualitative Information
JEL Codes: G23, G24, L26
I. Introduction
It is broadly documented in extant literature that information asymmetry in security
offering increases the cost of raising capital (Leuz and Verrecchia, 2000; Hughes, Liu and Liu,
2007; Armstrong, Core, Taylor, and Verrecchia, 2011; Shroff, Sun, White, and Zhang 2013).
Consequently, it is in firms’ interest to voluntarily disclose information to the public (Botosan
1997; Lang and Lundholm, 2000; Korajczyk, Lucas and McDonald, 2015), even when the
disclosures are viewed as only partially credible (Pownall and Waymire 1989). In a broader
context, voluntary disclosure increases the liquidity of equity markets (Diamond and Verrecchia
1991; Welker 1995), which has a significant impact on asset prices (Acharya and Pedersen,
2005).
Security offering does not have to take place in well-established exchanges, in which
high listing requirements and expensive floatation costs turn down many young and small
businesses. Instead, equity crowdfunding provides a new channel for entrepreneurs to raise
money directly from a group of investors through online security offering. Transactions between
entrepreneurs and investors often take place in an equity crowdfunding platform.
Contrary to traditional stock exchanges, equity crowdfunding platforms adopt much more
flexible listing requirements and do not involve underwriters in administering security offering.
Entrepreneurs post detailed information regarding firm, management, product market and
financial records etc. on a crowdfunding platform; investors then evaluate the quality of a firm
based on the information and decide whether to invest. Entrepreneurs do not disclose information
through platforms after crowdfunding campaigns. However, they must report business conditions,
financial records and crowdfunding results to Securities and Exchange Commission (SEC). So
far, platforms do not support a secondary market for equity crowdfunding investors.
The flexible listing requirements and slack corporate disclosure policy make equity
crowdfunding especially exposed to information asymmetry. Admittedly, inferior firms that
cannot obtain investment from Angel Investors or Venture Capitalists may take advantage of the
low listing requirements and choose equity crowdfunding as the last resort. This adverse
selection problem reduces the overall quality of crowdfunding projects and is detrimental to
investor confidence in the equity crowdfunding market.
As the value of crowdfunding firms is relatively stable during fundraising process, the
information asymmetry problem disappears when investors learn about quality (Kirmani and Rao,
2000). The important questions are: what approaches do high-quality firms take to reveal quality
to investors? Does it pay off for an average or low-quality firm to signal quality? Can low-
quality firms cheat investors through window dressing?
The answers to the research questions reside in the information entrepreneurs provided.
In general, information about startups seeking crowdfunding can be divided into two parts: the
quantitative part, which includes financial statements in previous years, management educational
level and years of relevant industry experience, estimated product market size and the qualitative
part, which includes entrepreneurs’ introduction on business model, competitive strategy,
product market, drivers and barriers for product/service adoption and business milestones. While
quantitative business information is data driven and often directly verifiable, qualitative
information reflects entrepreneurs’ self-evaluation on the business and can be hardly verified.
In regards to interpretation of descriptive or qualitative firm information, extant studies
have documented that positive and negative language has a substantial influence on how
information is processed (Davis, Piger, and Sedor, 2012). In particular, the tone entrepreneurs
and managers use in describing business conditions is informative and has impact on investor
responses. For instance, managers’ use of optimistic language increases litigation risk (Rogers,
Buskirk, and Zechman, 2011); the tone of earnings announcement related conference calls
reflects current and future firm performance and is significantly influenced by a manager-
specific tendency to be optimistic or pessimistic (Davis, Ge, Matsumoto, and Zhang, 2015); the
tone of Form S-1 significantly affect investors’ ability to value IPO (Loughran and McDonald,
2013); the level of pessimistic language in the Management’s Discussion and Analysis included
with firms’ SEC 10-Q or 10-K filings is predictive of firm performance in future quarters (Davis
and Sweet, 2012).
Extant studies on qualitative information focus exclusively on public corporations, which
are larger in size, receive broader media coverage, have higher liquidity in equity markets, and
face less information asymmetry problem than private small businesses. As market liquidity
drops and the information asymmetry problem escalates, how investors respond to qualitative
business information is an important field remains unexplored.
In this paper, I use empirical evidence in the field of small business financing to
demonstrate that, in addition to quantitative firm information, qualitative business descriptions
also has a significant impact on equity crowdfunding outcomes. In particular, more detailed
qualitative business introduction is associated with higher chance of funding success and larger
amount of capital raised; and this effect is especially obvious for high quality firms. Furthermore,
excessive use of promotional language, measured by percentage of positive words used in
entrepreneurs’ qualitative introduction on the business, negatively affects fundraising outcomes.
I argue that the entrepreneurs’ qualitative business introduction signals unobservable firm quality
and entrepreneur credibility.
This paper is directly related to studies on equity crowdfunding mechanisms. In particular,
retaining equity and providing more detailed information about risks are effective signals on
corporate quality and have a strong impact on fundraising success (Ahlers, Cumming, Günther,
and Schuweizer, 2015); increasing goal size is negatively related with probability of fundraising
success (Vulkan, Astebro, and Sierra, 2016); illustration videos facilitate fundraising while
increased fundraising duration decreases the chances of crowdfunding success (Mollick, 2014).
My study contributes to the existing literature from three aspects: first, it reveals the
impact of qualitative business information on mitigating the information asymmetry between
small business entrepreneurs and investors; second, it demonstrates that crowdfunding investors
do not blindly accept signals on business quality from entrepreneurs and that seemingly costless
entrepreneur self-promotion, if not supported by quantitative business information, reduces the
probability of fundraising success; third, it examines crowdfunding investors’ preference and
offers strategic fundraising guidance to small businesses seeking external financing.
II. Institutional Background
Equity crowdfunding is not available to small business entrepreneurs seeking external
financing until very recent years. As a new financing channel alternative to the traditional Angel
and Venture Capital investment, equity crowdfunding provides small business entrepreneurs an
opportunity to sell a specified amount of equity or bond-like shares to a pool of investors via an
online platform (Ahlers, Cumming, Günther, and Schuweizer, 2015). This democratic financing
method enables crowdfunding entrepreneurs to adopt more flexible strategies in promoting their
businesses to an unspecific crowd than to targeted institutional investors. Above all, what
entrepreneurs need is investment from a proportion of potential investors, not endorsement from
everyone in the crowd. Consequently, it is not impossible for entrepreneurs to strategically target
a subgroup of investors or even window dress a business to cheat unsophisticated investors.
The dynamics of equity crowdfunding can be analyzed starting from two simple
questions: which firms can be listed on an equity crowdfunding platform? Which investors can
participate in equity crowdfunding?
The first question pertains to the listing requirement of equity crowdfunding platforms. In
general, SEC requires equity crowdfunding platforms to apply due diligence to prevent
fraudulent fundraising1 but does not provide explicit guidelines on detailed procedures platforms
should follow to ensure the genuineness of crowdfunding projects. In practice, platforms have
ample flexibility in determining which projects to list. Platform managers usually discuss with
entrepreneurs to assess proposed business plans before publicly releasing project information on
a platform2. Some platforms conduct entrepreneur background verification, personal meeting or
on-site visits, financial or credit checks, cross checks from social media connections, monitoring
account activities and requesting third party certificates or proof on applied projects for investor
protection (Cumming and Zhang, 2016). By and large, any firm free of potential fraud concerns
discerned by a fundraising platform may launch an equity crowdfunding campaign. However,
equity crowdfunding platforms may strategically impose additional project listing criteria to
differentiate themselves from market competitors. Firms seeking equity crowdfunding are
different in industry classification and in developing and production stages—from a rudimentary
idea conception to mature products or service with proven marketing records. The fundraising
targets of equity crowdfunding firms usually range from several hundred thousand dollars to
1 SEC open meeting fact sheet, October 20, 2015.
2 Based on SEC requirements, information on an equity crowdfunding project has to be publicly disclosed on a registered crowdfunding platform
for a minimum 21 days before security offering.
over a million. Depending on platforms’ policy, entrepreneurs may keep the entire amount raised,
regardless of whether they meet fundraising goal, or keep nothing unless the fundraising goal is
achieved (Cumming, Leboeuf and Schwienbacher, 2017).
The second question pertains to investor qualification in the evolving legislative
environment. Equity crowdfunding is subject to high levels of regulation (Heminway and
Hoffman, 2010) and was not generally permitted until very recent years (Cumming and Johan,
2013; Mollick, 2014). Early on, to address small business entrepreneurs’ increasing need of
alternative financing channels, a few state securities regulators exempted equity crowdfunding
within their respective jurisdictions: Kansas first legalized equity crowdfunding under the Invest
Kansas Exemption (IKE) Act, which became effective in August 2011; a similar Invest Georgia
Exemption Act became effective in December 2011, followed by Idaho (July, 2012)3, Michigan
(December 2013), and Wisconsin (November 2013). By 2015, equity crowdfunding was
legalized in 17 states4 while 14 states
5 have proposed rules or introduced bills on equity
crowdfunding exemptions (Bohliqa, 2015). However, intrastate equity crowdfunding exemptions
have natural limitations: small businesses have to be headquartered and investors have to be
residents in the state. Different investment limits were also imposed on crowdfunding investors:
from $100 for Maryland to $10,000 for Georgia, Michigan and Wisconsin, while Washington
permitted an investment of 10% of annual income for investors with annual income over
$100,000. Most states only granted accredited investors, measured by net worth or annual
income, access to equity crowdfunding. The exception was Maine, in which issuers are permitted
to sell equity to unaccredited investors. After various intrastate equity crowdfunding exemptions,
3 Idaho’s exemption on equity crowdfunding is based on administrative order
4 KS, GA, ID, MI, DC, TX, VT, AL, WA, WI, IN, MD, MR, MA, OR, TN, VA
5 NM, MS, CT, FL, HI, IA, KY, MN, NE, NH, NJ, NC, UT, WV
on October 30, 2015, SEC officially announced the nationwide adoption of Title III of Jumpstart
Our Business Startups (JOBS) Act, under which ordinary investors are permitted to invest in
equity crowdfunding projects. Title III of the JOBS Act became effective on May 16, 2016.
III. Hypotheses
Equity investors are profit-driven. They participate in highly risky and illiquid
crowdfunding market in the hope of sharing sizeable long-term profit with the next Google or
Facebook. Nevertheless, failure in entrepreneurship is pervasive, especially for early stage
startups (Mcgrath, 1999). Equity crowdfunding investors, realizing the risk of investing in
startups, are selective in crowdfunding projects and sensitive to information from entrepreneurs.
A part of firm information overlooked by extant studies is entrepreneurs’ qualitative
descriptions on business model, competitive strategy, product market, drivers and barriers for
product/service adoption and business milestones. Introductions on startup name, industry,
location, products function and service types are compulsory identification information and not
deemed as qualitative business information. Qualitative business information is important
because it provides another channel for investors to evaluate a firm thus narrows the information
gap with investors. Although providing qualitative business information is not compulsory,
entrepreneurs who do not seize the opportunity to communicate with potential investors send out
an undesirable signal on startup quality.
One critical issue is measuring readability of qualitative business description. Inspired by
Loughran and Mcdonald (2011, 2014), who report that the frequency of negative words used in
public firms’ 10-K documents reflects management insight and predicts future stock
performance, I use total number of words in entrepreneurs’ qualitative business introductions to
evaluate the level of disclosure on qualitative startup characteristics and percentage of
promotional words used in qualitative business introduction to measure the extent of
entrepreneurs’ overstatement on startup quality.
Whether qualitative business information facilitates equity crowdfunding is an important
question remains unanswered.
On the one hand, qualitative business information not only reveals startups’ strategic
framework, advantages over market competitors, growth potential, expected risk and operational
records, but also signals entrepreneurs’ business vision and professionalism. This information is
complementary to quantitative records and enables potential investors to better assess startup
quality through classical methods such as SWOT (strength, weakness, opportunities and threats)
analysis. Investors thus face less uncertainty in decision making and invest more in a startup.
On the other hand, qualitative business information can be cheap talk or biased self-
appraisal, as it is not directly verifiable. Moreover, most entrepreneurs release only favorable
startup information in fundraising campaigns. Consequently, the impact of qualitative
information on crowdfunding is compromised, resulting in little change in investor behavior.
Hypothesis 1: More detailed disclosure of qualitative business information increases
the chance of funding success and leads to larger amount of capital raised.
In the context of equity crowdfunding, startup quality can be reflected by the percentage
of fundraising plan completed: a startup is of high quality when percentage of fundraising target
completed ranks at top quantile; low quality if ranks at bottom quantile. If qualitative business
information is effective in mitigating the information gap between entrepreneurs and
crowdfunding investors, high quality startups should take advantage of the opportunity to signal
value to potential investors, as once quality is revealed, they become more attractive investment
targets. However, whether it is in low quality firms’ interest to disclose qualitative business
information remains unclear.
A main attribute of low quality startups is insufficient, not necessarily inferior accounting
records. Many firms seeking equity crowdfunding are freshly incorporated and face a natural
obstacle in time frame to signal value to potential investors. In addition, due to lack of financial
support, entrepreneurs who possess innovative products or service ideas and competitive
operational strategies often cannot find opportunities to implement their business plans.
Furthermore, as it takes time for a firm to obtain market traction, build customer loyalty and
reach economy of scale, even a firm with inferior accounting records in the concurrent period
could still have strong growth potential in the long run. However, as growth potential cannot be
accurately reflected by quantitative business records, these startups face a severe disadvantage in
equity crowdfunding and consequently have low chance to achieve fundraising targets. Under
this circumstance, qualitative business introduction provides low quality startups a valuable
opportunity to elaborate their business vision, strategic plans, and competitive advantages,
resulting in better investor recognition and higher chance to achieve fundraising targets. From
another perspective, as high quality startups reveal their value to investors through qualitative
business introduction, failure to provide qualitative information sends potential investors a clear
bad signal on quality. In either case, it is in low quality startups’ interest to disclose qualitative
business information to potential investors.
Alternatively, low quality startups with weak quantitative records may actually benefit
from the information asymmetry in equity crowdfunding at the expense of high quality startups,
provided that investors conjecture unobservable quality to be at average level. For a low quality
startup, disclosing qualitative business information may unintentionally expose its weaknesses to
potential investors, resulting in increased chance of failure in equity offering. Consequently, low
quality startups may not necessarily benefit from releasing qualitative business information.
Hypothesis 2: It is in low quality firms’ interest to disclose qualitative business
information to potential investors.
If disclosing qualitative business information is beneficial to both high quality and low
quality startups, whether and how the according benefit vary over startup quality is a question
worthy examine.
On the one hand, due to concerns on entrepreneurs’ qualification and credibility,
investors do not blindly accept qualitative introductions on startup quality. Above all, stating a
business plan is much easier than implementing it. This effect is especially obvious for
entrepreneurs without relevant industry experience. In response to these concerns, investors refer
to quantitative startup records when evaluating the feasibility of qualitative information: ceteris
paribus, solid financial and marketing records signal high feasibility and therefore enhance the
impact of qualitative information while weak quantitative records reduce the impact of
qualitative information on fundraising. As strong quantitative records are naturally favored by
investors, the corresponding businesses have high probability to achieve fundraising targets and
are recognized as high quality startups. These high quality startups with strong quantitative
records thus benefit more from disclosure of qualitative information than low quality startups.
On the other hand, if qualitative business information facilitates fundraising through
mitigating the information gap between entrepreneurs and crowdfunding investors, then startups
exhibiting the largest information gap, which attributes to insufficient or weak quantitative
business records, should benefit most from releasing qualitative business information. By
comparison, startups with complete and strong quantitative business records are naturally more
favorable to investors and have higher chances to achieve funding targets and to raise more
capital from investors, resulting in smaller benefit from releasing qualitative business
information. In this case, low quality startups should benefit more from disclosing qualitative
business information than high quality firms.
Hypothesis 3: The impact of qualitative business information on crowdfunding
outcomes vary across startup qualities: high quality startups benefit more from
disclosing qualitative business information than low quality startups.
Since entrepreneurs have ample flexibility in descripting businesses, they are inclined to
overstate quality to facilitate fundraising. Inevitably, entrepreneurs’ sentiment and opinions are
interwoven with factual statements in qualitative business information. In particular, promotional
words6 are used to convey business value and advertise funding campaigns to potential investors.
Correspondingly, investors realize the biasness of qualitative business information and face
uncertainties in entrepreneurs’ credibility. Investors are thus cautious about the promotional
language used in crowdfunding campaigns: for a given length of qualitative business information,
imprudent usage of promotional words indicates exaggerated startup quality and compromised
entrepreneurs’ credibility. In this regard, excessive use of promotional words in qualitative
business information negatively affects crowdfunding outcomes.
Alternatively, promotional words in qualitative business information reflect entrepreneurs’
confidence, optimism and business insight. Psychological research indicates that successful
6 A word is considered promotional if it reflects entrepreneurs’ subjective opinion on the business, management team, product or service, intends
to bring favorable effects to fundraising and cannot be directly verified.
entrepreneurs have a psychological profile different from the less successful ones (Veciana 2007).
Consequently, these valuable personal traits, reflected by positive words in qualitative business
information, should be favored by crowdfunding investors. It is also possible that investors are
not sophisticated enough to disentangle false quality signals from normal business promotions
and therefore do not associate excessive promotional words with entrepreneurs’ credibility. In
either case, extensive use of promotional words in qualitative business information does not
overshadow crowdfunding outcomes.
Hypothesis 4: Excessive use of promotional language negatively impact
crowdfunding outcomes.
IV. Data
The data were provided by EquityNet7, an online platform focuses exclusively on equity
crowdfunding. The sample data disclose fundraising activities for 6870 startups, covering a
period from January 2007 to November 2016. In the sample, 5273 or 76.8% of startups are
incorporated in United States; the rest are international firms. No startup launched more than one
fundraising campaign over the sample period.
Startups in the sample cover 18 different industry sectors8; among which, the most
represented sectors are manufacturing (16.4% of startups), information and cultural industries
(11.6%), professional, scientific and technical services (11.1%) and retail trade (10.9%).9 The
amounts of capital startups seek ranges from $15,000 to $5,000,000. Over the sample period,
7 https://www.equitynet.com/
8 Industry classification is based on the first two digits of North American Industry Classification System (NAICS) code.
9 Other industry sector representations are: real estate and rental and leasing (7.7%), health care and social assistance (5.1%), other services
except public administration (4.8%), arts, entertainment and recreation (4.6%), administrative and support, waste management and remediation
services (4.2%), finance and insurance (4.1%), construction (4.0%), transportation and warehousing (3.4%), accommodation and food services
(3.1%), educational services (2.6%), wholesale trade (2.0%), utilities(1.8%), mining, quarrying, and oil and gas extraction (1.4%) and agriculture, forestry, fishing and hunting (0.8%).
2,497 or 36.4% of startups have fully achieved their fundraising targets. On average, a startup
raises $300,000 through equity crowdfunding.
A representative startup10
in the data is incorporated in United States 1 year prior to
crowdfunding campaign and has around $100,000 annual revenue before equity offering. It is
operated by 2 male managers who hold undergraduate or graduate degrees but have no related
industry experience. The startup reports an estimated product/service market size of 3.6 billion
USD and wants to raise $400,000 USD through offering around 70% of common shares.
However, the startup neither provides estimated firm value nor indicates planned exit channel11
to potential investors. In addition, it does not provide entrepreneur photos or adopt videos in
business introduction. The representative startup cannot achieve its fundraising target.
Table 1 reveals general characteristics of crowdfunding activities in the sample.
Table 1 is here
Table 1 shows that a representative startup uses 182 words to introduce business model,
competitive strategy, product market, drivers and barriers for product/service adoption and
business milestones to potential investors. Qualitative business information is widely provided
by crowdfunding entrepreneurs to reveal quality and reduce information gap with potential
investors.
Figure 1 exhibits a general view on disclosure of qualitative business information among
startups seeking equity crowdfunding.
10
A representative startup exhibits median values for variables included in summary statistic table. 11
Startup entrepreneurs may indicate their planned exit channels for potential investors. The following investment exit channels are specified by
entrepreneurs seeking external financing from crowd: business merger, acquisition, management buy-out, IPO, issuing debt (short-term, long-term and convertible) to buy back equity, and secondary private offering.
Figure 1 is here
Figure 1 shows that, in regards to qualitative business introduction, 23.9% of startups use
10 or fewer words; 15.2% of startups use between 11 words and 100 words; 22.8% of startups
use between 101 words and 300 words; 15.1% of startups use between 301 words and 500 words;
17.4% of startups use between 501 words and 1000 words; 5.6% of startups use more than 1000
words. The impact of qualitative business information on crowdfunding outcome is briefly
illustrated in Figure 2.
Figure 2 is here
Figure 2 shows that the length of qualitative business introduction is positively associated
with percentage of fundraising plan completed. Specifically, the average number of introductory
words used by startups that achieved 0% of funding target is 0, 10% of funding target is 4.8, 20%
of funding target is 20.8, 30% of funding target is 70.6, 40% of funding target is 126.0, 50% of
funding target is 177.2, 60% of funding target is 263.1, 70% of funding target is 367.3, 80% of
funding target is 363.6, 90% of funding target is 413.7, and 100% of funding target is 566.3.
Among qualitative introductory words used by the representative startup, 25 words are
identified as promotional words because they reflect entrepreneur’s positive self-appraisal on the
business. These words are adopted to highlight startup quality, despite that the underlying
content is often biased and not verifiable.
Table 2 shows the most frequently used promotional words in equity crowdfunding
campaigns. A complete vocabulary list of promotional words is provided in appendix.
Table 2 is here
Figure 3 illustrates the distribution on usage of promotional words among startups
seeking equity crowdfunding.
Figure 3 is here
Figure 3 shows that for the general sample, 24.2% of startups do not use any promotional
language. In addition, promotional words account for 6.4% of total words used in qualitative
business introduction on average and 7.4% for a representative startup. Within the subsample of
startup firms that provide qualitative business information, promotional words account for 8.4%
of total words used on average and 8.3% for a representative startup.
Table 3 reveals the general correlations among variables describing equity crowdfunding
activities. Consistent with analyses in prior studies, both percentage of fundraising plan
completed and amount of capital raised are positively associated with entrepreneurs’ educational
level and relevant industry experience, estimated product market size, firm revenue, equity
retention ratio, estimated startup value, availability of entrepreneurs’ photo and video illustration.
These correlations are statistically significant at 1% level.
Table 3 is here
Table 3 also shows that, contrary to conclusions in extant literature, both percentage of
fundraising plan completed and amount of capital raised are also positively associated with
number of managers and amount of capital seeking, indicating that in general, large startups are
more likely to achieve better equity crowdfunding outcomes than small startups. One possible
explanation is that large startups that seek more capital from investors are better prepared in
equity crowdfunding campaigns through disclosing more detailed business information; their
better communication with potential investors overcomes size disadvantage. In addition,
percentage of R&D expense, difficult level of staffing, and located in U.S. are positively
correlated with better equity crowdfunding outcomes, indicating investors’ preference on
business innovation, sophistication of human capital and domestic startups.
It is also worth noting that length of qualitative business introduction is positively
associated with higher fundraising plan completed and larger amount of capital raised, indicating
that qualitative information facilitates fundraising. Surprisingly, frequency of promotional words
used is positively correlated with crowdfunding outcomes. One possible explanation is that
frequency of promotional words is associated with length of business introduction; the positive
correlation may capture the impact of detailed qualitative introduction on crowdfunding
outcomes. Further detailed variable relationships are presented in the correlation matrix.
The interpretation on variable correlations is not conclusive, as it does not control for
possible confounding factors. The following multivariate analyses investigate equity
crowdfunding activities in detail.
V. Multivariate Analyses
This section examines the marginal impacts of qualitative business introduction and
usage of promotional language on equity crowdfunding outcomes. Analyses on percentage of
fundraising plan completed are presented first, followed by analyses on amount of capital raised
from crowdfunding campaigns.
An important concern on the validity of analyses pertains to potential endogeneity
problem between qualitative information and fundraising outcome. On the one hand, fundraising
outcome reflects investors’ decisions, which are based on their interpretation on startup quality;
and qualitative startup information reveals quality. Therefore, qualitative information determines
fundraising outcome. On the other hand, entrepreneurs may learn investors’ preference and
intentionally inflate the content of qualitative information to increase chance of successful
fundraising. It is also possible that unobserved factors may impact both entrepreneurs’ delivery
of qualitative information and crowdfunding outcome. To address this potential endogeneity
problem, Two-Stage Least Squares (2SLS) regression is used.
For analyses about the impact of qualitative startup information on fundraising outcome,
the basic model specifications follow:
(1) Length of Qualitative Introduction= Constant + β1*Percentage of Females in
Management Team + β2 * Controls+ Residuals
(2) Fundraising Outcome = Constant + β1 * Predicted Length of Qualitative Introduction +
β2 * Controls+ Residuals
In the first stage regression, length of qualitative introduction is regressed against
percentage of females in management team and control variables. Percentage of females in
management team is used as instrument variable because psychology studies (for instance, Aries
1987, Archer 2001) indicate that men and women have different habits in expression and
communication. As firms seeking equity crowdfunding are generally small in size, qualitative
information provided to potential investors is likely to be reviewed by everyone in the
management team; consequently, percentage of female managers impact the length of qualitative
introduction. In addition, since gender is not associated with managerial ability, percentage of
female managers has no impact on fundraising outcome.
In the second stage regression, predicated length of qualitative information is used to
explain fundraising outcome.
The regression controls for the following five categories of variables: first, operational
characteristics, including startup age, annual revenue, availability of estimated business value
prior to crowdfunding campaign, R&D expense, and domestic versus foreign dummy variable;
second, human resource characteristics, including number of managers, industry experience and
educational level of management team and difficult level of staffing; third, market potential, i.e.
estimated product market size; fourth, equity offering conditions, including equity retention ratio,
fundraising target and availability of planned investment exit channel; fifth, campaign
promotional tactics, including availability of video illustration and entrepreneur photos.
The regression is estimated as a cross-sectional model clustering on startup industry
classification and equity crowdfunding campaign initiation month. Observations are clustered
into industries based on the first two digits of NAICS code to account for industry specific
characteristics. Standard errors are also clustered by time, based on campaign initiation month,
because macroeconomic factors and systematic risk could lead to correlations among
crowdfunding outcomes.
Similarly, for analyses about the impact for percentage of promotional words used in
qualitative introduction on fundraising outcome, the basic model specifications follow:
(3) Percentage of Promotional Words in Qualitative Introduction=Constant+ β1 * Percentage
of Females in Management Team + β2 * Controls+ Residuals
(4) Fundraising Outcome = Constant + β1 * Predicted Percentage of Promotional Words in
Qualitative Introduction + β2 * Controls+ Residuals
Table 4 present regression results on percentage of fundraising plan completed. Model 1-
4 focus on the impact of qualitative business information whereas Model 5-8 analyze the impact
of promotional language.
Table 4 is here
Table 4 shows that more detailed qualitative business introduction leads to higher
percentage of fundraising plan completed. On average, a one standard deviation increase or an
increase of 362 words in length of qualitative introduction increases percentage of fundraising
plan completed by 19.0% (Model 4)—23.1% (Model 1); this effect is statistically significant at 1%
level. In addition, excessive usage of promotional language results in lower percentage of
fundraising plan completed. On average, a one standard deviation or 4% increase in percentage
of promotional words used in qualitative business introduction reduces percentage of fundraising
plan completed by 5.2% (Model 7)—6.1% (Model 8); this effect is statistically significant at 1%
level. The results are consistent with Hypothesis 1 and 4.
Table 4 also shows that in respect of human recourse, number of managers has a negative
impact on percentage of fundraising plan completed, but the economic significance is small: on
average, a one standard deviation increase in manager number, or an increase of 2 managers
reduces the percentage of fundraising plan completed by 1.3% (Model 8)—3.8% (Model 1); this
effect is statistically significant at 10% in Model 8, at 5% in Model 6 and Model 7 and at 1% in
Model 1—5. Moreover, richer industry experience and higher educational level in the
management team lead to better funded results: on average, a one standard deviation increase or
an increase of 9.6 years spent in relevant industry leads to 2.8% (Model 6)—4.5% (Model 1)
increase in percentage of fundraising plan completed, a one scale increase in educational level
increases the percentage of plan completed by 3.6% (Model 8)—4.9% (Model 1); both effects
are statistically significant at 1% level. Difficult level of staffing, which indicates employee
sophistication, also has a positive impact on percentage of plan completed: a one scale increase
in difficult level of staffing increases percentage of fundraising plan completed by 3.5% (Model
8)—4.7% (Model 4); this effect is statistically significant at 1% level. Estimated market size and
startup revenue also exhibit positive impact on percentage of plan completed, but the economic
significances are small: a ten times increase in estimated product market size or annual revenue
increases percentage of fundraising plan completed by 4.2%—6.2% or 4.3%—5.7% respectively;
the effects are statistically significant at 1% level. The insensitivity of fundraising outcome with
respect to estimated market size reveals that investors are cautious about entrepreneurs’
estimation on general market potential and focus more on the competitiveness of a startup. In
addition, the positive impact of startup revenue is mitigated by investors’ preference on smaller
startups, which report less revenue but have lower fundraising targets.
Table 4 points out that whether a startup indicates planned exit channel to potential
investors is of critical importance on percentage of fundraising plan completed. On average, a
startup indicates planned investment exit channel to potential investors is 35.8% (Model 8)—
43.8% (Model 1) more likely to achieve fundraising target than a comparable startup that does
not; this effect is statistically significant at 1% level. Over the sample period, no startup fulfills
its fundraising plan without indicating planned investment exit channel. Although firms seeking
equity crowdfunding face high uncertainties at early development stage and their planned
investment exit channels may never be realized, setting up a target investment exit channel
signals the direction of strategic business movement and reveals entrepreneurs’ attitude on
fiduciary duty in asset management. In addition, whether a startup indicates estimated firm value
prior to a crowdfunding campaign also impacts fundraising outcome: on average, a startup with
estimated firm value available to potential investors is 3.9% (Model 2)—6.1% (Model 8) more
likely to achieve fundraising target than a comparable startup without estimated firm value; this
effect is statistically significant at 1% level. Further analyses reveal that in respect to percentage
of fundraising plan completed, investors do not exhibit preferences on particular investment exit
channels and are insensitive to the specific amount of estimated startup value; the according
results are not reported for conciseness.
Table 4 further reveals that investors have strong domestic preference: on average, a
domestic startup is 11.9% (Model 4)—14.8% (Model 7) more likely to achieve fundraising target
than a comparable foreign startup. In addition, providing entrepreneur photos facilitates
fundraising: on average, a one standard deviation increase in average number of photos per
manager increases the percentage of fundraising plan completed by 2.4% (Model 3)—2.8%
(Model 7). Investors also exhibit preference on younger startups, but the according economic
significance is small: on average, a one year increase in startup age reduces percentage of
fundraising plan completed by 0.4% (Model 8)—0.5% (Model 4). These effects are all
statistically significant at 1% level. Startup R&D expense does not exhibit statistically significant
impact on percentage of fundraising plan completed.
Table 5 uses quantile regressions to evaluate impact variations regarding qualitative
business information and promotional language on percentage of fundraising plan completed.
Table 5 is here
Table 5 shows that qualitative business information has positive impact on fundraising
across different levels of startups quality, consistent with Hypothesis 2. In addition, the impact of
qualitative business information on fundraising increases monotonically with respect to startup
quality. Specifically, at 20th
percentile, a one standard deviation increase in length of qualitative
business information increases the percentage of fundraising plan completed by 6.4%; at 40th
percentile, 15.1%; at 60th
percentile, 20.6%; at 80th
percentile, 25.6%. These effects are
statistically significant at 1% level.
Table 5 also reveals that the impact of promotional language on fundraising intensifies
with improvement in startup quality. Specifically, at 20th
percentile, a one standard deviation
increase in percentage of promotional words used reduces the percentage of fundraising plan
completed by 2.8%; at 40th
percentile, 4.9%; at 60th
percentile, 6.8%; at 80th
percentile, 7.7%.
This effect is statistically significant at 5% level in Model 5 and at 1% level in Models 6—8.
Hausman specification tests indicate statistically significant differences in coefficients for
qualitative business information and percentage of promotional words used in Models1—4 and
Model 5—8, respectively, consistent with Hypothesis 3.
Table 6 shows that qualitative business information and promotional language have
positive and negative impacts, respectively, on amount of capital raised from equity
crowdfunding campaigns. On average, a one standard deviation increase or an increase of 362
words in length of qualitative business information increases amount of capital raised by a
significant 89.2% (Model 1)—139.49% (Model 3); a one standard deviation or 4% increase in
percentage of promotional words used in qualitative business introduction reduces amount of
capital raised by 3.8% (Model 5)—5.2% (Model 6); these effects are consistent with Hypothesis
1 and 4 and statistically significant at 1% level.
Table 6 is here
Table 6 also shows that, consistent with the results in Table 4, managerial industry
experience, educational level, availability of investment exit channel, estimated product/service
market size, and firm revenue have positive impact on amount of capital raised. Specifically, on
average, a one standard deviation increase in years of relevant industry experience increases the
amount of capital raised by 30.1% (Model 6)—35.8% (Model 5); a one scale increase in
managerial educational level increases the amount of capital raised by 4.9% (Model 8)—6.7%
(Model 5); a startup with planned exit channel available raises a significant 218.4% (Model 1)—
244.2% (Model 4) more capital than a comparable startup without indication of planned
investment exit channel; a 10 times increase in estimated product market size increases the
amount of capital raised by 10.1% (Model 8)—12.9% (Model 1); a 10 times increase in firm
revenue increases the amount of capital raised by 21.2% (Model 6)—25.2% (Model 4). These
effects are statistically significant at 1% level.
Table 6, however, exhibits different results than Table 4 in that number of managers is
positively associated with amount of capital raised. On average, a one standard deviation
increase in manager number, or an increase of 2 managers increases the amount of capital raised
by 34.0% (Model 7)—43.8% (Model 5). In addition, a 10 times increase in estimated startup
value increases the amount of capital raised by 18.8% (Model 4)—33.7% (Model 1). R&D
expense also has positive impact on amount of capital raised: a one standard deviation increase
in percentage of R&D expense to annual revenue increases the amount of capital raised by 3.3%
(Model 8)—4.3% (Model 2). Startup age also has positive impact on amount of capital raised: a
one year increase in startup age on average increases the amount of capital raised by 3.3%
(Model 3)—4.1% (Model 4). These effects are at least statistically significant at 5% level.
Table 7 analyzes the subsample of startups that achieved 100% of fundraising targets for
the impact of qualitative business information and usage of promotional language on amount of
capital raised. Regressions in Table 7 control for variables exhibiting statistical significance in
Table 6. However, fundraising target is not included as a control variable: for fully funded
startups, amount of capital raised equals amount of capital seeking. Overfunding is not observed
over the sample horizon. All fully funded startups have indicated planned exit channels.
Table 7 is here
Table 7 further reveals the impact of qualitative business information and percentage of
promotional words on amount of capital raised. On average, for fully funded startups, a one
standard deviation increase in length of qualitative business introduction, or an increase of 80
words, increases the amount of capital raised by 28.6% (Model 3)—42.0% (Model 1); a one
standard deviation or 4% increase in percentage of promotional words used in qualitative
business introduction reduces the amount of capital raised by 3.7% (Model 5)—3.9% (Model 4).
These effects are statistically significant at 1% level. The analyses results for the subsample are
consistent with Hypothesis 1 and 4.
Table 7 also shows that, number of managers, managerial industry experience, estimated
startup value, estimated product market size and firm value all have positive impact on amount
of capital raised from equity crowdfunding for the subsample of fully funded startups; the effects
are at least statistically significant at 5% level. Specific marginal effects for the subsample
analyses are not reported for conciseness.
VI. Robustness Checks
It is worth noting that potential measurement error can bias multivariate analyses.
Specifically, the length and percentage measurement on qualitative business information and
usage of promotional language may not accurately reflect the amount of information delivered to
crowdfunding investors. Entrepreneurs’ expression habits and writing skills also influence the
content of communication: depending on sentence structure, 200 words may convey the same
information as 250 words. To account for differences in sentence structure, I sort the length of
qualitative business information and percentage of promotional words into 4 quantiles and use
scale indicator from 1 to 4 in lieu of exact number of words or percentages to measure the level
of detail in qualitative business information and extent of usage of promotional language. The
results are presented in appendix as robustness checks.
Appendix Table 2—Appendix Table 5 show that the impacts of qualitative business
information and usage of promotional language on equity crowdfunding outcomes are robust to
measurements differences.
To further examine the endogeneity concerns regarding the impact of qualitative
information on fundraising outcome, I divide the observations into two groups based on whether
the length of qualitative information is above group median, or the existence of “treatment
effect”. Startups with more qualitative information are classified into the “treatment group”; less
qualitative information, “control group”. The startups in the treatment group are then matched
with startups in the control group based on the control variables used in the multivariate analyses.
The matching process is based on propensity score matching (PSM), nearest-neighbor matching
(NNM) or inverse probability weighted regression adjustment (IPWRA) method. The impact of
qualitative information on fundraising is then revealed through the average treatment effect in
population and the average treatment effect on the treated group.
The matched sample analyses in Appendix Table 6 suggest that more qualitative startup
information leads to higher percentage of fundraising plan completed and larger amount of
capital raised from equity crowdfunding campaigns, consistent with results in main tables.
In addition, actual length of qualitative startup information and observed frequency of
promotional words are used as explanatory variables in analyses to check the robustness of using
predicted values in the 2SLS Analyses. The results are strongly consistent with that in main
tables. For conciseness, relevant regressions are not reported.
VII. Discussion
The success of equity crowdfunding depends critically on effective communication
between startup entrepreneurs and potential investors: entrepreneurs reveal quality, potential
investors learn quality, and investment decisions are made based on quality interpretation.
Empirical data indicate that quantitative startup information is at the center of quality
revealing: startup revenue, managers’ educational degree and years of experience in relevant
industry, and level of employee sophistication all have significant positive impact on
crowdfunding outcome. However, the data-driven information has its natural limitations.
Important startup characteristics, such as business model feasibility, competitive advantage,
strategic development plan, and barriers and risks for expansion cannot be inferred from the
quantitative startup records. In this regard, qualitative startup introduction serves as a
supplemental channel for entrepreneurs to communicate with investors. Empirical data reveal
that more detailed disclosure of qualitative information leads to higher investor recognition and
better fundraising results. It even pays off for low quality startups to disclosure qualitative
business information to potential investors. From investors’ perspective, the worst quality signal
is no signal at all.
Empirical data show that in the quality learning process, crowdfunding investors do not
blindly accept qualitative business information: quantitative startup records are commonly used
to evaluate the reliability of entrepreneurs’ qualitative business introductions. The marginal
benefit of disclosing qualitative information on fundraising outcome increases with improvement
of quantitative startup records. Consequently, high quality startups benefit more from disclosure
of qualitative information than low quality startups. In this regard, qualitative startup information
catalyzes the impact of quantitative business records on equity crowdfunding.
In addition, the quality interpretation process reveals investors’ preferences.
Empirical data further show that investors have strong preference on domestic, small and
young firms. The investors’ domestic preference reflects more than simple home bias, but
prudent concerns on asset protection due to differences in legislative environments and
efficiency of law enforcement. Smaller firms are favored because they offer investors a higher
proportion of equity for a given amount of capital investment, have less team conflicts among
managers, and adapt faster to evolving business environment than larger firms. In addition, older
firms face significantly higher performance requirements than younger firms, as they have more
time and opportunity to obtain market traction and receive investments from alternative channels.
Surprisingly, empirical data suggest limited explanatory power of firm revenue on
crowdfunding outcomes, indicating that investors do not emphasize a firm’s current productivity.
This low explanatory power is however consistent with the study of Crozet, Head, and Mayer
(2012), who claim that the principal difference between firms is quality, not physical
productivity. In addition, the impacts of estimated startup value and product market size on
fundraising are not as strong as expected, indicating that investors are cautious about the risks
embedded in long-term cash flow. Conversely, the availability of planned investment exit
channel is vital in determining fundraising success, indicating investors’ strong concern about
the opportunity of capital withdrawal.
Empirical data also reveal that promotional language is not favored by crowdfunding
investors. Given that qualitative business introduction disclosures inside information, the content
of disclosure is often not verifiable. In this case, entrepreneurs’ positive self-appraisal on the
business is recognized by investors as window dressing rather than objective evaluation.
Crowdfunding investors need factual basis and rationale explanations supporting qualitative
business information more than biased opinions from fundraisers. An isolated check mark given
by startup entrepreneurs amplifies, not mitigates the information gap with potential investors and
is detrimental to fundraising outcome.
In a broad sense, effectiveness of communication has profound impact on equity
crowdfunding industry.
Mutual trust is the foundation of effective communication. Equilibrium is reached when
entrepreneurs honestly introduce startup characteristics and crowdfunding investors trust the
information received. The equilibrium breaks down when fraudulent fundraising projects are
listed on an equity crowdfunding platform and investors do not have the ability to disentangle
inferior projects from valuable ones.
Equity crowdfunding fraud can occur in different formats. In the most extreme case,
“entrepreneurs” raise money and disappear. More generally, inflating quantitative business
records, overstating strategic partnerships, fabricating business milestones, and exaggerating
entrepreneurial experience are all examples of fraud in equity crowdfunding. Battling against
fraud is critical in protecting investors’ interest and in ensuring the healthy development of
crowdfunding industry. Under this circumstance, equity crowdfunding platforms are in the
forefront of detecting and preventing fraud: they are the gateways of fundraising, have access to
posted startup information and can compare one crowdfunding project with others to monitor
suspicious activities. Whether and how crowdfunding platforms conduct due diligence in
assuring the genuineness of crowdfunding projects has profound influence on fundraising
outcomes. One possible approach for platforms is setting up a fraud record and sharing it with
other crowdfunding platforms: once fraud is detected, the responsible project initiator is
blacklisted and banned by all platforms from engaging in future crowdfunding activities. As cost
of fraud escalates, we should observe increases in volume and efficiency in equity crowdfunding.
The sophistication of crowdfunding investors depends on participation requirements and
supportive educational environment. Prior to Title III of JOBS Act, equity crowdfunding is
commonly restricted to high-net-worth investors. As wealth is in general positively associated
with knowledge and investment skills, the average investor’s quality evaluation ability drops
when ordinary investors are granted entry permission. This reduced investor sophistication
provides new opportunities for crowdfunding fraud, elevates the necessity for investor protection
and calls for more thorough investor education. Currently, although SEC requires equity
crowdfunding platforms to provide educational materials to investors, exactly what platforms do
to inform investors and the impact of investor education is an important topic remains unstudied.
Finally, on entrepreneurs’ side, quality signaling is essential for crowdfunding success. In
crowdfunding campaigns, videos and pictures are commonly used to attract investors’ attention.
However, attention is not equivalent to recognition, which is based on investors’ interpretation
on startup quality. Due to time constraint, quantitative startup records can be hardly improved
during fundraising period; nevertheless, professional expression in qualitative business
introduction is a valuable skill that can be quickly trained and implemented, leading to more
efficient quality signaling and better fundraising outcome.
VIII. Conclusion
Information asymmetry is widely documented in security offering. The information gap
between entrepreneurs and investors enlarges when security offering qualification drops, market
liquidity descends and media coverage retreats. To facilitate security offering, entrepreneurs
have to mitigate the information gap through quality revealing.
This paper assesses quality signaling and receiving between entrepreneurs and investors
in equity crowdfunding and analyzes the impact of effective communication on crowdfunding.
The data suggest that startups seeking equity crowdfunding disclose both quantitative and
qualitative business information to investors. Specifically, quantitative information is data-driven:
accounting records, managers’ educational level and years of industry experience, estimated
product market size and startup value are all quantitative information. Qualitative information is
description based, covering introductions on business model, competitive strategy, product
market, drivers and barriers for product/service adoption and business milestones.
The data show that qualitative business introduction is effective in mitigating the
information asymmetry between entrepreneurs and investors and strongly impacts equity
crowdfunding outcome. In general, more detailed disclosure of qualitative information leads to
better fundraising outcome. In addition, it pays off for startups of all quality levels to disclose
qualitative information to crowdfunding investors.
The data reveal that investors do not blindly accept qualitative information. They cross
check startups’ quantitative background to evaluate the reliability of qualitative business
introduction. Consequently, high quality startups benefit more from disclosing qualitative
information than low quality startups. Investors also refrain from supporting startups that use
excessive promotional language in business introductions. The ineffective promotional language
is recognized by investors as window dressing due to lack of adequate factual support.
Overall, empirical evidence leads to the conclusion that when revealing firm quality,
equity crowdfunding entrepreneurs effectively disclose qualitative information in addition to
quantitative records to crowdfunding investors. In general, qualitative information is quality
revealing because it reveals business characteristics not observable from quantitative records and
provides investors an alternative channel to evaluate the underlying firm.
References
Acharya, V.V., Pedersen, L.H. 2005. Asset pricing with liquidity risk. Journal of Financial
Economics, Volume 77, Issue 2, August 2005, pp. 375-410
Ahlers, G.K.C., Cumming, D., Günther, C., Schweizer, D. 2015. Signaling in Equity
Crowdfunding. Entrepreneurship Theory and Practice, 2015, Volume 39, Issue 4, July 2015
pp. 955–980
Archer, D., Silver, J., Nonverbal Workshop (Firm), & Berkeley Media. (2001). Gender &
Communication: Male-Female Differences in Language & Nonverbal Behavior.
Aries, E. (1987). Gender and communication. In P. Shaver & C. Hendrick (Eds.), Review of
personality and social psychology, Vol. 7. Sex and gender (pp. 149-176). Thousand Oaks,
CA: Sage Publications.
Armstrong, C.S., Core, J.E., Taylor, D.J., Verrecchia, R.E. 2011. When Does Information
Asymmetry Affect the Cost of Capital? Journal of Accounting Research, Vol. 49 No. 1
March 2011
Bohliqa, A. 2015. A Study of the U.S. Intrastate Crowdfunding Exemptions. Portland
International Conference on Management of Engineering and Technology (PICMET), pp.
961-967
Botosan, C.A. 1997. Disclosure Level and the Cost of Equity Capital. The Accounting Review,
volume 72, No. 3, July 1997, pp. 323-349
Crozet, M., Head, K., Mayer, T. 2012. Quality Sorting and Trade: Firm-level Evidence for
French Wine, Review of Economic Studies, Volume 79, Issue 2, pp. 609-644
Cumming D., Johan, S. 2013. Demand Driven Securities Regulation: Evidence from
Crowdfunding. Venture Capital: An International Journal of Entrepreneurial Finance,
Volume 15, September 2013, pp. 361-379
Cumming D., Leboeuf, G., Schwienbacher, A., 2017. Crowdfunding Models: Keep-It-All vs.
All-Or-Nothing. Working Paper, https://papers.ssrn.com/sol3/Papers.cfm?abstract_id=2447567
Cumming D., Zhang, Y., 2016. Are Crowdfunding Platforms Active and Effective
Intermediaries? Working Paper, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2882026
Davis, A.K., Ge, W., Matsumoto, D., Zhang, J.L. 2015. The effect of manager-specific optimism
on the tone of earnings conference calls. Review of Accounting Studies, June 2015, Volume
20, Issue 2, pp. 639–673
Davis, A.K., Piger, J. M., Sedor, L.M. 2012. Beyond the Numbers: Measuring the Information
Content of Earnings Press Release Language. Contemporary Accounting Research, Volume
29, Issue 3, September 2012, pp. 845-868
Davis, A.K., Sweet, I.T. 2012. Managers’ Use of Language across Alternative Disclosure Outlets:
Earnings Press Releases versus MD&A. Contemporary Accounting Research, Volume 29,
Issue 3, September 2012, pp. 804-837
Diamond, D.W., Verrecchia, R.E. 1991. Disclosure, Liquidity, and the Cost of Capital. The
Journal of Finance, Vol. 46, No. 4, Sep., 1991, pp. 1325-1359
Healy, P.M., Palepu, K.G. 2001. Information asymmetry, corporate disclosure, and the capital
markets: A review of the empirical disclosure literature. Journal of Accounting and
Economics, Volume 31, Issues 1–3, September 2001, pp. 405-440
Heminway, J., Hoffman, S. 2010. Proceed at your peril: crowdfunding and the securities act of
1933. Tennessee Law Review, Volume 78 (2010), pp. 879-972
Hughes, J.S., Liu, J. Liu, J. 2007. Information Asymmetry, Diversification, and Cost of Capital
The Accounting Review, Vol. 82, No. 3, May 2007, pp. 705-729
Kirmani, A., Rao, R.A. 2000. No Pain, No Gain: A Critical Review of the Literature on
Signaling Unobservable Product Quality. Journal of Marketing, April 2000, Volume 64, No.
2, pp. 66-79
Korajczyk, R.A., Lucas, D.J., McDonald, R.L., 1991. The Effect of Information Releases on the
Pricing and Timing of Equity Issues. The Review of Financial Studies, Volume 4 Issue 4,
October 1991, pp. 685-708
Lambert, R., Leuz, C., Verrecchia, R.E. 2007. Accounting Information, Disclosure, and the Cost
of Capital. Journal of Accounting Research, Volume 45, Issue 2, May 2007, Pages 385–420
Lang, M.H., Lundholm, R.J., 2000. Voluntary Disclosure and Equity Offerings: Reducing
Information Asymmetry or Hyping the Stock? Contemporary Accounting Research, Volume
17, Issue 4, winter 2000, pp. 623-62
Leuz, C., Verrecchia, R.E. 2000. The Economic Consequences of Increased Disclosure. Journal
of Accounting Research, Vol. 38, 2000, pp. 91-124
Loughran, T., Mcdonald, B. 2011. When Is a Liability Not a Liability? Textual Analysis,
Dictionaries, and 10-Ks. The Journal of Finance, Volume 66, Issue 1, February 2011, pp.
35–65
Loughran, T., McDonald, B. 2013. IPO first-day returns, offer price revisions, volatility, and
form S-1 language. Journal of Financial Economics, 2013, vol. 109, issue 2, pp. 307-326
Loughran, T., Mcdonald, B. 2014. Measuring Readability in Financial Disclosures. The Journal
of Finance, Volume 69, Issue 4, August 2014, pp. 1643–1671
McGrath, R. G. 1999. Falling Forward: Real Options Reasoning and Entrepreneurial Failure.
The Academy of Management Review, Volume 24, No. 1, January 1999, pp.13-30
Mollick, E., 2014. The dynamics of crowdfunding: An exploratory study. Journal of Business
Venturing, Volume 29, Issue 1, January 2014, pp. 1-16
Pownall, G., Waymire. G. 1989. Voluntary Disclosure Credibility and Securities Prices:
Evidence from Management Earnings Forecasts. Journal of Accounting Research, Volume
27, No. 2, autumn 1989, pp. 227-245
Rogers, J.L., Buskirk, A.V., Zechman, S.L.C. 2011. Disclosure Tone and Shareholder Litigation.
The Accounting Review, November 2011, Vol. 86, No. 6, pp. 2155-2183
Shroff, N., Sun, A.X., White, H.D., W. Zhang 2013. Voluntary Disclosure and Information
Asymmetry: Evidence from the 2005 Securities Offering Reform. Journal of Accounting
Research, Volume 51, Issue 5, December 2013, pp. 1299–1345
Veciana, J. M. 2007. Entrepreneurship as a Scientific Research Programme. Entrepreneurship:
Concepts, Theory and Perspective, January 2007, pp 23-71
Vulkana, N., Åstebro, T., Sierra, M.F. 2016. Equity crowdfunding: A new phenomena. Journal
of Business Venturing Insights, Volume 5, June 2016, pp. 37-49
Welker, M. 1995. Disclosure Policy, Information Asymmetry, and Liquidity in Equity Markets.
Contemporary Accounting Research, Volume 11, Issue 2, spring 1995, pp. 801–827
TABLE 1: Variable Definitions and Summary Statistics
Table 1 defines the variables in the data set and provides summary statistics for equity crowdfunding activities in EquityNet between January2007and November 2016.
Variable Name Definition Min Mean Max S.D. Obs
Percent of Fundraising Plan Completed Percentage of fundraising target achieved: 1 stands for 100%. 0.0000 0.6305 1.0000 0.3457 6870
Amount of Capital Raised (log transformed) Total amount of capital raised in U.S. dollar through an equity
crowdfunding campaign. Data is log transformed. 6.8024 12.6204 14.9141 2.1583 6594
Length of Qualitative Business Description Number of words used in qualitative business description. 1 unit
stands for 100 words. 0.0000 3.0483 17.5300 3.6239 6870
Percent of Promotional Words Used in Business Description
Number of promotional words used divided by total number of words used in business description
0.0000 0.07489 0.2886 0.04120 6870
Amount of Capital Seeking (log transformed) Amount of capital a startup seeks from investors. Data is log
transformed. 9.6158 13.2636 15.4250 1.2254 6870
Number of Managers Number of managers in a startup 1.0000 2.5265 12.0000 2.1359 6870
Percentage of Female Managers Percentage of female managers in management team 0.0000 0.2751 1.0000 0.3149 6870
Average Industry Experience Average years in corresponding industry for startup managers 0.0000 6.9444 40.0000 9.6280 6870
Average Education Level
Average educational level for startup managers in scale: high school 1, some undergraduate/associates 2, Completed Undergraduate 3,
Some Graduate 4, Completed Graduate 5, Some Doctorate 6,
Completed Doctorate 7.
1.0000 3.2683 7.0000 1.4582 6870
Estimated Product Market Size (log transformed) Estimated product market size in U.S. dollars. Data is log
transformed. 15.4249 25.5675 29.5301 3.2646 6870
Firm Revenue (log transformed) Startup revenue in U.S. dollars in the year prior to crowdfunding campaign. Data is log transformed. If no revenue is reported, log
transformed value is treated as 0.
0.0000 12.8610 14.4885 2.6490 6870
Equity Retention Ratio Percentage of common shares a startup will retain after crowdfunding 0.01458 0.3638 0.9126 0.2479 6870
Percentage of R&D Expense Startup R&D expense over revenue in the year prior to crowdfunding
campaign. If no revenue is reported, median R&D expense ratio of
corresponding industry in the sample data-set is used.
0.0000 0.1003 0.5594 0.1306 6870
Difficult Level of Staffing The scale is 1 - 7, where 7 indicate staffing is very difficult. 1.0000 5.1307 7.0000 1.6694 6870
Planed Exit Channel Available? (Yes=1; No=0) Does a startup indicate planned exit channel in crowdfunding
campaign? Yes=1; No=0 0.0000 0.4773 1.0000 0.4995 6870
Pre-funding Startup Value Available? (Yes=1; No=0) Does a startup indicate estimated firm value prior to crowdfunding
campaign? Yes=1; No=0 0.0000 0.4242 1.0000 0.4943 6870
Estimated Startup Value (log transformed) Estimated firm value prior to crowdfunding campaign in U.S. dollars.
Data is log transformed. 10.8198 14.7859 16.1181 1.4220 6870
U.S. Firm? (Yes=1; No=0) Does a startup incorporated in U.S.? Yes=1; No=0 0.0000 0.7675 1.0000 0.4224 6870
Startup Age Number of complete years from startup incorporation to
crowdfunding campaign 0.0000 2.9237 18.0000 4.7709 6870
Average Number of Photos per Manager Total number of management team photos divided by number of
managers 0.0000 0.3487 1.0000 0.4131 6870
Video Used in Fundraising Campaign? (Yes=1; No=0) Is there a video embedded in crowdfunding campaign? Yes=1; No=0 0.0000 0.2256 1.0000 0.4180 6870
TABLE 2: Most Frequently Used Promotional Words
Table 2 presents the top 50 most frequently used promotional words in equity crowdfunding campaigns. Promotional words are used by startup entrepreneurs to emphasize startup quality and attract investor attention.
Cumulative Frequency Cumulative Frequency
new 6727 advantage 1233
more 5824 improve/improved/improvement 1227
grow/growing 5732 best 1150
high 4930 significant 1144
add value 4595 established 1120
large/largely/larger/largest 3700 rise 1110
increase/increased/increasing 3695 effective 1098
well 3482 good 1073
first 3424 cutting edge 1058
most 3141 big/bigger/biggest 1049
value/valuable 2850 efficient/efficiency 1043
proper 2228 strong 1029
quality 2114 innovate/innovation/innovative 996
experienced 1920 much 996
expand/expansion 1718 safe 984
wide 1707 early 981
key 1666 fast 966
competitive 1617 better 943
success/successful 1604 international 939
unique/uniquely 1592 quick 770
lead 1587 super/superior/superlative 761
top 1473 smart 692
great 1465 broad 670
profession/professional/professionalism 1424 advance 627
global 1419 expert 614
TABLE 3: Correlation Matrix
Table 3 presents the correlations among variables of interest. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Percent
Plan
Completed
Amount
Cap.
Raised
Length of
Introduction
Percent of
Promotional
Words
Amount
Cap.
Seeking
No.
Managers
Ave.
Industry
Experience
Ave.
Edu.
Level
Market
Size
Firm
Revenue
Equity
Retention
Ratio
Per. R&D
Expense
Difficult
Staffing
Level
Exit
Channel
Available
Pre-
funding
Startup
Value
Available
Est.
Startup
Value
U.S. Firm Startup
Age
Ave.
Photo
Number
Video
Used
Percent Plan Completed. 1
Amount Cap. Raised 0.793*** 1
Length of Introduction 0.616*** 0.547*** 1
Percent of Promotional
Words 0.494*** 0.463*** 0.645*** 1
Amount Capital Seeking 0.386*** 0.372*** 0.541*** 0.457*** 1
Number of Managers 0.544*** 0.539*** 0.565*** 0.508*** 0.535*** 1
Average Industry Experience 0.596*** 0.554*** 0.463*** 0.400*** 0.550*** 0.484*** 1
Average Educational Level 0.550*** 0.476*** 0.502*** 0.422*** 0.672*** 0.547*** 0.612*** 1
Market Size 0.503*** 0.670*** 0.614*** 0.503*** 0.664*** 0.524*** 0.530*** 0.359*** 1
Firm Revenue 0.445*** 0.531*** 0.595*** 0.520*** 0.724*** 0.556*** 0.531*** 0.540*** 0.669*** 1
Equity Retention Ratio 0.497*** 0.514*** 0.391*** 0.356*** 0.513*** 0.388*** 0.367*** 0.429*** 0.446*** 0.519*** 1
Per. R&D Expense 0.264*** 0.252*** 0.209*** 0.186*** 0.250*** 0.218*** 0.197*** 0.259*** 0.236*** 0.269*** 0.198*** 1
Difficult. Staffing Level 0.207*** 0.398*** 0.348*** 0.238*** 0.396*** 0.264*** 0.303*** 0.384*** 0.456*** 0.361*** 0.216*** 0.102*** 1
Exit Channel Available 0.642*** 0.591*** 0.532*** 0.455*** 0.490*** 0.521*** 0.557*** 0.697*** 0.681*** 0.537*** 0.553*** 0.270*** 0.414*** 1
Pre-funding Value Available 0.578*** 0.429*** 0.528*** 0.460*** 0.626*** 0.514*** 0.541*** 0.650*** 0.652*** 0.520*** 0.624*** 0.270*** 0.366*** 0.880*** 1
Est. Startup Value 0.354*** 0.332*** 0.554*** 0.490*** 0.528*** 0.545*** 0.549*** 0.646*** 0.656*** 0.437*** 0.401*** 0.276*** 0.350*** 0.349*** 0.464*** 1
U.S. Firm 0.383*** 0.242*** 0.224*** 0.156*** 0.242*** 0.184*** 0.215*** 0.225*** 0.263*** 0.218*** 0.143*** 0.0685*** 0.342*** 0.244*** 0.216*** 0.213*** 1
Startup Age -
0.0974***
-
0.0381** -0.00643 0.00934
-
0.0374** 0.0139 0.0851*** -0.0139
-
0.0391** -0.0257* -0.00230 0.00574 -0.119***
-
0.0394** -0.0231 -0.00268 -0.100*** 1
Ave. Photo Number 0.454*** 0.421*** 0.331*** 0.280*** 0.417*** 0.245*** 0.372*** 0.418*** 0.398*** 0.364*** 0.264*** 0.117*** 0.241*** 0.413*** 0.376*** 0.376*** 0.120*** -0.0166 1
Video Used 0.197*** 0.217*** 0.259*** 0.234*** 0.214*** 0.223*** 0.160*** 0.175*** 0.207*** 0.212*** 0.196*** 0.0846*** 0.0886*** 0.191*** 0.212*** 0.235*** 0.0431*** 0.0537*** 0.243*** 1
TABLE 4: General Analyses on Percentage of Fundraising Plan Completed
Table 4 presents cross sectional regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. The dependent variable is the percentage of fundraising plan completed for a crowdfunding project. Standard errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at
the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Predicted Percentage of Promotional Words Used
-0.815*** -0.809*** -0.787*** -0.924***
(-3.26) (-3.55) (-3.14) (-2.87)
Predicted Length of Qualitative Business Description 0.0402*** 0.0385*** 0.0377*** 0.0330***
(2.78) (3.25) (3.34) (4.11)
Number of Managers -0.0112*** -0.00835*** -0.00826*** -0.00659*** -0.00573*** -0.00510** -0.00422** -0.00376*
(-3.66) (-3.29) (-3.16) (-3.47) (-3.09) (-2.21) (-1.97) (-1.88)
Average Industry Experience 0.00295*** 0.00268*** 0.00283*** 0.00264*** 0.00198*** 0.00183*** 0.00231*** 0.00211***
(4.75) (4.28) (4.39) (5.04) (4.77) (4.35) (4.24) (4.38)
Average Educational Level 0.0308*** 0.0255*** 0.0269*** 0.0247*** 0.0251*** 0.0246*** 0.0230*** 0.0228***
(8.95) (8.13) (7.56) (10.17) (13.04) (13.21) (12.55) (11.52)
Estimated Product Market Size (log transformed) 0.0162*** 0.0171*** 0.0166*** 0.0143*** 0.0136*** 0.0133*** 0.0128*** 0.0114***
(5.84) (5.68) (5.63) (4.81) (8.95) (8.86) (9.02) (8.43)
Firm Revenue (log transformed) 0.0143*** 0.0122*** 0.0135*** 0.0118*** 0.0139*** 0.0156*** 0.0140*** 0.0129***
(10.54) (9.37) (10.19) (10.27) (9.22) (9.60) (9.53) (8.42)
Planned Exit Channel Available? (Yes=1; No=0) 0.276*** 0.269*** 0.253*** 0.241*** 0.254*** 0.237*** 0.228*** 0.226***
(15.35) (15.29) (15.76) (14.63) (15.44) (14.71) (15.19) (15.33)
Pre-funding Startup Value Available? (Yes=1; No=0)
0.0247*** 0.0285** 0.0316***
0.0346*** 0.0359*** 0.0385***
(3.28) (2.39) (2.74)
(3.31) (3.27) (3.25)
Difficult Level of Staffing
0.0243*** 0.0248*** 0.0295***
0.0251*** 0.0239*** 0.0220***
(4.69) (5.33) (4.87)
(5.62) (6.16) (5.85)
U.S. Firm? (Yes=1; No=0)
0.0765*** 0.0753***
0.0931*** 0.0885***
(3.44) (3.67)
(3.46) (3.81)
Average Number of Photos per Manager
0.0359*** 0.0384***
0.0422*** 0.0405***
(4.52) (4.70)
(3.68) (3.93)
Startup Age
-0.00316***
-0.00283***
(-3.08)
(-3.24)
Percentage of R&D Expense
0.0384
0.0460
(1.39)
(1.55)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 0.134*** 0.113*** 0.328** 0.197*** 0.158*** 0.0957*** 0.109*** 0.124***
(3.24) (2.72) (2.28) (3.54) (4.13) (3.17) (4.08) (3.62)
Number of Observations 6870 6870 6870 6870 6870 6870 6870 6870
Number of Clusters (industry classification) 20 20 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.230 0.241 0.252 0.261 0.235 0.242 0.250 0.258
TABLE 5: Quantile Analyses on Percentage of Fundraising Plan Completed
Table 5 presents quantile regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. Observations are divided into quantiles
based on predicted startup quality specified in equation 5. The dependent variable is the percentage of fundraising plan completed for a crowdfunding project. Standard errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1
(Quantile 0.20) Model 2
(Quantile 0.40) Model 3
(Quantile 0.60) Model 4
(Quantile 0.80) Model 5
(Quantile 0.20) Model 6
(Quantile 0.40) Model 7
(Quantile 0.60) Model 8
(Quantile 0.80)
Predicted Percentage of Promotional Words Used
-0.425** -0.736*** -1.023*** -1.158***
(-2.16) (-4.65) (-3.42) (-4.14)
Predicted Length of Qualitative Business Description 0.0112*** 0.0263*** 0.0359*** 0.0446***
(4.12) (6.28) (6.14) (5.25)
Number of Managers -0.00194 -0.00208** -0.00277* -0.00338* -0.00143 -0.00189** -0.00246* -0.00293*
(-1.27) (-2.23) (-1.74) (-1.69) (-1.03) (-2.33) (-1.82) (-1.70)
Average Industry Experience 0.00246*** 0.00229*** 0.00207*** 0.00218*** 0.00261*** 0.00230*** 0.00215*** 0.00197***
(3.16) (4.02) (2.94) (3.65) (2.66) (4.27) (4.36) (3.72)
Average Educational Level 0.0166*** 0.0185*** 0.0173*** 0.0241*** 0.0199*** 0.0218*** 0.0191*** 0.0226***
(7.16) (9.38) (9.59) (11.64) (8.28) (9.46) (10.73) (11.28)
Estimated Product Market Size (log transformed) 0.0135*** 0.0142*** 0.0124*** 0.0138*** 0.0137*** 0.0144*** 0.0119*** 0.0133***
(4.68) (5.96) (5.75) (5.15) (6.75) (7.24) (6.43) (5.94)
Firm Revenue (log transformed) 0.0221*** 0.0147*** 0.00894*** 0.00689*** 0.0209*** 0.0132*** 0.00875*** 0.00568***
(8.26) (8.15) (6.17) (4.02) (8.73) (8.28) (6.29) (3.97)
Planned Exit Channel Available? (Yes=1; No=0) 0.213*** 0.254*** 0.316*** 0.285*** 0.204*** 0.262*** 0.306*** 0.277***
(10.49) (12.17) (12.74) (11.46) (11.23) (12.32) (12.53) (11.19)
Pre-funding Startup Value Available? (Yes=1; No=0) 0.0652*** 0.0417** 0.0256** 0.00713 0.0734*** 0.0365** 0.0196** 0.00855
(2.71) (2.25) (2.19) (1.54) (3.25) (2.34) (2.38) (1.21)
Difficult Level of Staffing 0.0174*** 0.0250*** 0.0237*** 0.0268*** 0.0188*** 0.0239*** 0.0247*** 0.0254***
(3.47) (4.50) (3.91) (3.35) (3.68) (4.85) (4.29) (4.38)
Average Number of Photos per Manager 0.0372*** 0.0296*** 0.0277*** 0.0289*** 0.0364*** 0.0275*** 0.0283*** 0.0332***
(3.84) (4.26) (4.13) (3.62) (3.49) (3.45) (3.21) (2.87)
U.S. Firm? (Yes=1; No=0) 0.0935*** 0.0864*** 0.0643*** 0.0430** 0.0942*** 0.0795*** 0.0535*** 0.0410***
(2.86) (3.67) (3.21) (2.19) (3.18) (3.39) (3.22) (2.73)
Startup Age -0.00572*** -0.00238*** -0.00195*** -0.00224*** -0.00543*** -0.00281*** -0.00204*** -0.00218***
(-3.37) (-2.81) (-3.25) (-3.07) (-2.95) (-3.20) (-3.19) (-2.97)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 0.141** 0.132*** 0.156*** 0.229** 0.0765*** 0.194*** 0.105*** 0.143**
(2.25) (3.18) (3.25) (2.76) (2.69) (3.17) (2.85) (2.13)
Number of Observations 6870 6870 6870 6870 6870 6870 6870 6870
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.243 0.231 0.223 0.235 0.238 0.234 0.241 0.237
TABLE 6: General Analyses on Amount of Capital Raised
Table 6 presents cross sectional regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. The dependent variable is the amount of capital raised from an equity crowdfunding campaign after log transformation. Due to log transformation, startups that receive no investment from crowdfunding are excluded from analyses.
Standard errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Predicted Percentage of Promotional Words Used
-0.937** -1.274*** -1.262*** -1.245***
(-2.49) (-2.80) (-3.26) (-3.25)
Predicted Length of Qualitative Business Description 0.176*** 0.193*** 0.241*** 0.216***
(6.74) (6.05) (6.18) (5.77)
Number of Managers 0.164*** 0.149*** 0.140*** 0.151*** 0.170*** 0.156*** 0.137*** 0.148***
(5.60) (4.49) (4.11) (3.78) (6.21) (4.05) (3.77) (3.85)
Average Industry Experience 0.0311*** 0.0285*** 0.0294*** 0.0297*** 0.0318*** 0.0273*** 0.0286*** 0.0288***
(6.33) (5.45) (5.03) (4.64) (5.93) (4.57) (4.67) (3.92)
Average Educational Level 0.0623*** 0.0576*** 0.0604*** 0.0512*** 0.0644*** 0.0550*** 0.0582*** 0.0476***
(5.16) (5.25) (4.37) (3.84) (4.92) (5.09) (4.26) (3.50)
Planned Exit Channel Available? (Yes=1; No=0) 1.158*** 1.165*** 1.169*** 1.236*** 1.174*** 1.182*** 1.164*** 1.179***
(9.17) (8.92) (8.67) (9.05) (10.21) (9.85) (9.17) (9.94)
Estimated Startup Value (log transformed) 0.126*** 0.112*** 0.0875*** 0.0747*** 0.124*** 0.108*** 0.0793*** 0.0802***
(4.32) (3.84) (3.90) (2.68) (3.94) (3.74) (2.93) (2.66)
Estimated Product Market Size (log transformed) 0.0526*** 0.0483*** 0.0515*** 0.0437*** 0.0495*** 0.0457*** 0.0423*** 0.0418***
(8.29) (7.44) (8.45) (8.07) (9.38) (8.30) (7.59) (6.68)
Firm Revenue (log transformed)
0.0952*** 0.0934*** 0.0975***
0.0836*** 0.0839*** 0.0843***
(9.13) (9.09) (8.71)
(9.38) (8.46) (8.44)
Percentage of R&D Expense
0.323*** 0.306** 0.287**
0.274*** 0.266** 0.249**
(2.72) (2.33) (2.15)
(2.63) (2.39) (2.04)
U.S. Firm? (Yes=1; No=0)
0.0167* 0.0168
0.0867 0.0886
(1.76) (1.57)
(0.96) (1.22)
Startup Age
0.0328*** 0.0403***
0.0375*** 0.0384***
(3.68) (3.70)
(4.28) (3.87)
Difficult Level of Staffing
-0.00246
-0.00324
(-1.44)
(-1.51)
Average Number of Photos per Manager
0.0678
0.0736
(0.43)
(0.31)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 1.097*** 1.023** 0.976*** 1.014*** 1.134*** 1.146*** 1.139*** 1.135**
(3.28) (2.52) (2.71) (3.45) (3.03) (2.92) (3.15) (2.27)
Number of Observations 6594 6594 6594 6594 6594 6594 6594 6594
Number of Clusters (industry classification) 20 20 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.187 0.203 0.225 0.226 0.166 0.194 0.223 0.224
TABLE 7: Subsample Analyses on Amount of Capital Raised
Table 7 presents cross sectional regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. Only fully funded equity
crowdfunding projects are included in the analyses. The dependent variable is the amount of capital raised from an equity crowdfunding campaign after log transformation. Standard errors are clustered
by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Predicted Percentage of Promotional Words Used
-1.148*** -1.087*** -1.091***
(-3.18) (-2.79) (-2.64)
Predicted Length of Qualitative Business Description 0.436*** 0.385*** 0.313***
(16.23) (15.38) (13.49)
Number of Managers 0.131*** 0.135*** 0.129*** 0.178*** 0.164*** 0.152***
(6.21) (5.82) (5.40) (5.43) (4.86) (4.57)
Average Industry Experience 0.0538*** 0.0415*** 0.0384*** 0.0426*** 0.0361*** 0.0356**
(3.74) (3.01) (2.83) (2.87) (2.66) (2.35)
Average Educational Level 0.0692* 0.0458 0.0263 0.0578 0.0494 0.0195
(1.77) (1.14) (0.65) (0.76) (0.58) (0.61)
Estimated Startup Value (log transformed) 0.196*** 0.185*** 0.182*** 0.188*** 0.184*** 0.173***
(8.43) (7.68) (7.49) (7.54) (7.08) (6.89)
Estimated Product Market Size (log transformed) 0.146*** 0.132*** 0.131*** 0.135*** 0.133*** 0.128***
(10.18) (11.30) (11.43) (11.82) (10.46) (11.24)
Firm Revenue (log transformed) 0.148*** 0.137*** 0.135*** 0.172*** 0.176*** 0.163***
(2.71) (2.84) (2.73) (2.79) (2.88) (2.67)
Percentage of R&D Expense
-0.404 -0.277
-0.165 -0.149
(-1.03) (-0.85)
(-0.93) (-1.24)
Startup Age
0.00363
0.00758
(0.35)
(0.47)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? No No No No No No
Constant 1.128*** 1.173*** 1.105*** 1.736*** 1.714*** 1.703***
(2.96) (3.23) (3.19) (2.88) (3.85) (3.60)
Number of Observations 2497 2497 2497 2497 2497 2497
Number of Clusters (industry classification) 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115
R2 0.145 0.147 0.147 0.133 0.134 0.135
FIGURE 1
Figure 1 illustrates the general conditions regarding adopting qualitative business introduction among startups seeking equity crowdfunding.
05
10
15
20
25
% S
am
ple
Firm
s
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Number of Words Used in Startup Introduction
(Measured in 100 Words; Winsorized at 1% and 99%)
Figure 1: Adoption of Qualitative Business Introduction
FIGURE 2
Figure 2 illustrates the impact of adopting qualitative business introduction on percentage of fundraising plan completed in equity crowdfunding. Each dot reflects the condition for one startup; the fitted
line shows the evolving trend based on median value of each startup group classified by percentage of fundraising plan completed.
05
10
15
20
Nu
mb
er
of W
ord
s U
se
d in S
tart
up
In
trod
uctio
n (
in 1
00 W
ord
s)
0 10 20 30 40 50 60 70 80 90 100
Percent of Fundraising Plan Completed
Figure 2: Impact of Qualitative Business Information on Crowdfunding Outcome
FIGURE 3
Figure 3 illustrates the general conditions regarding adopting promotional words in qualitative business introduction among startups seeking equity crowdfunding.
05
10
15
20
25
% S
am
ple
Firm
s
0 .05 .1 .15 .2 .25
Propotion of Promotional Words in Startup Introduction
Figure 3: Adoption of Promotional Words in Qualitative Business Introduction
APPENDIX TABLE 1. Vocabulary List for Promotional Words Used in Equity Crowdfunding
This table presents the vocabulary list regarding promotional words used in equity crowdfunding. A word is considered promotional if it reflects entrepreneurs’ subjective opinion on the business, management team, product or service, intends to bring favorable effects to fundraising and cannot be directly verified. * indicates a negative prefix is used or the word is used to describe competitors.
abundant accelerate accurate active adaptable add
adept adequate advance advanced advantage affordable
aggressive aggressively amazing ambitious anywhere appreciative
approachable appropriate astonishing astute at all at least
attractive attuned benefit best better big
bigger biggest boast bold boost breakthrough
broad candid capable careful carefully catchy
champion classic clear comfortable committed competent
competitive complete comprehensive confident conscientious consistent
convenient cool creative credible critical customer-oriented
cutting edge daring decisive dedicated dependable detailed
detail-oriented determined diligent discreet diverse diversified
dominant dominate dramatic dramatically dynamic early
easier easily easy economical effective effectively
efficiency efficient energetic enhance enormous enough
enterprising enthusiastic entire established ethical ever
exceed excel excellent exceptional excite exciting
exclusive exemplary expand expansion expensive experienced
expert explore explosive extend extensive extensively
extraordinarily extraordinary extreme extremely failed* fair
APPENDIX TABLE 1 (continued)
fast favorable first flexibility flexible flourish
flourishing forefront friendly full gain game-changing
global good great grew grow grown
guarantee guaranteed happy hard-working harm* hazardous*
healthy heart help helpful high high-earning
honest huge hungry hyper ideal imaginative
important impress impressed impressive improve improved
improvement increase increased increasing independent indispensable
inferior* influential innovate innovation innovative insight
insightful intelligent international inviting key knowledgeable
large largely larger largest lead leading
less* limited long lasting love loyal major
majority many mass massive master mature
maximize meaningful meticulous moral more most
motivated much nation national new novel
numerous only open open-minded optimal optimize
optimized optimizing organic organized original outperform
outstanding overcome overwhelming passionate patient perceptive
perfect persevere persistent pioneer pioneered pleasant
APPENDIX TABLE 1 (continued)
poor* popular popularity positive powerful praise
precise predominant preferred premature premier premium
prepared primary proactive problem-solving productive professional
professionalism proficient profitable prolific promote proper
proprietary prosper proven prudent punctual qualified
quality quick radically raise rapid rational
reasonable recommended refine refined reinforce reliability
reliable remarkable respected revolutionary revolutionize rise
rising robust safe seamless sharply shortly
significant simple sizable skilled skillful smart
smarter sole solidifying soon stable starving*
strategic strategically streamline stress-free strong succeed
success successful sufficient super superior superlative
surpass surpassed surprising sustainability sweat tactful
talented tenacious thin* thorough time-saving top
tough tremendous trendy trumpet trustworthy unique
uniquely unparalleled upgraded uplift upsell upswing
useful valuable value variety various versatile
viability visionary vital well wide work-saving
world worldwide zealous
APPENDIX TABLE 2: General Analyses on Percentage of Fundraising Plan Completed (Robustness Check)
This table presents robustness checks on the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. The dependent variable is the percentage of fundraising plan completed for a crowdfunding project. Length of qualitative business introduction and percentage of promotional words used are sorted into 4 quantiles. Scale indicators from 1 to 4 in
lieu of exact number or percentage of words are used to measure the level of detail in qualitative business introduction and usage of promotional expression. Standard errors are clustered by startup
industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Predicted Percentage of Promotional Words Used in Scale
-0.677*** -0.701*** -0.854*** -0.833***
(-3.15) (-3.26) (-3.17) (-2.89)
Predicted Length of Qualitative Business Description in Scale 0.0368** 0.0352*** 0.0355*** 0.0290***
(2.44) (2.69) (2.74) (2.60)
Number of Managers -0.0123*** -0.0119*** -0.00965*** -0.00784*** -0.00481*** -0.00367** -0.00375** -0.00415*
(-3.27) (-3.45) (-3.23) (-2.66) (-2.84) (-2.47) (-2.13) (-1.92)
Average Industry Experience 0.00328*** 0.00311*** 0.00295*** 0.00272*** 0.00233*** 0.00204*** 0.00227*** 0.00215***
(5.06) (4.37) (4.28) (5.29) (4.35) (4.29) (4.16) (4.24)
Average Educational Level 0.0296*** 0.0251*** 0.0262*** 0.0238*** 0.0269*** 0.0234*** 0.0217*** 0.0221***
(7.64) (7.15) (8.03) (9.81) (10.22) (11.46) (10.28) (9.62)
Estimated Product Market Size (log transformed) 0.0174*** 0.0163*** 0.0167*** 0.0138*** 0.0150*** 0.0142*** 0.0126*** 0.0105***
(5.38) (5.46) (5.25) (4.42) (7.63) (6.85) (7.14) (6.13)
Firm Revenue (log transformed) 0.0159*** 0.0138*** 0.0122*** 0.0113*** 0.0146*** 0.0134*** 0.0128*** 0.0125***
(11.17) (10.03) (10.65) (10.24) (9.34) (9.57) (8.85) (8.69)
Planned Exit Channel Available? (Yes=1; No=0) 0.279*** 0.266*** 0.270*** 0.248*** 0.262*** 0.253*** 0.247*** 0.235***
(14.79) (14.29) (14.36) (14.28) (15.26) (14.77) (15.26) (14.48)
Pre-funding Startup Value Available? (Yes=1; No=0)
0.0255*** 0.0284*** 0.0303***
0.0367*** 0.0348*** 0.0376**
(3.66) (2.61) (2.74)
(3.23) (2.89) (2.48)
Difficult Level of Staffing
0.0414*** 0.0326*** 0.0297***
0.0242*** 0.0237*** 0.0236***
(3.74) (4.63) (4.15)
(4.60) (3.63) (4.25)
U.S. Firm? (Yes=1; No=0)
0.0815*** 0.0784***
0.0875*** 0.0760***
(3.22) (3.35)
(2.76) (2.81)
Average Number of Photos per Manager
0.0344*** 0.0369***
0.0403*** 0.0394***
(4.68) (4.31)
(3.82) (3.70)
Startup Age
-0.00278***
-0.00311***
(-2.96)
(-2.63)
Percentage of R&D Expense
0.0264
0.0302
(1.13)
(1.20)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 0.129** 0.108*** 0.233*** 0.220*** 0.142*** 0.104*** 0.113*** 0.117***
(2.46) (3.07) (3.51) (2.97) (3.21) (3.58) (3.19) (2.86)
Number of Observations 6870 6870 6870 6870 6870 6870 6870 6870
Number of Clusters (industry classification) 20 20 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.225 0.232 0.238 0.249 0.217 0.230 0.242 0.246
APPENDIX TABLE 3: Quantile Analyses on Percentage of Fundraising Plan Completed (Robustness Check)
This table presents robustness checks on quantile regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. The dependent variable is the percentage of fundraising plan completed for a crowdfunding project. Length of qualitative business introduction and percentage of promotional words used are sorted into 4 quantiles.
Scale indicators from 1 to 4 in lieu of exact number or percentage of words are used to measure the level of detail in qualitative business introduction and usage of promotional expression. Standard
errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1
(Quantile 0.20)
Model 2
(Quantile 0.40)
Model 3
(Quantile 0.60)
Model 4
(Quantile 0.80)
Model 5
(Quantile 0.20)
Model 6
(Quantile 0.40)
Model 7
(Quantile 0.60)
Model 8
(Quantile 0.80)
Predicted Percentage of Promotional Words Used in Scale
-0.467*** -0.663*** -0.852*** -1.116***
(-2.74) (-3.27) (-4.33) (-4.02)
Predicted Length of Qualitative Business Description in Scale 0.0143*** 0.0274*** 0.0395*** 0.0538***
(4.48) (7.24) (6.35) (5.74)
Number of Managers -0.00168 -0.00229** -0.00283** -0.00342 -0.00160 -0.00201** -0.00274* -0.00348**
(-0.66) (-2.26) (-1.99) (-1.54) (-0.85) (-2.46) (-1.79) (-1.97)
Average Industry Experience 0.00257*** 0.00230*** 0.00186*** 0.00211*** 0.00270*** 0.00238*** 0.00206*** 0.00184***
(3.28) (3.55) (4.04) (3.78) (3.44) (3.45) (3.28) (4.13)
Average Educational Level 0.0145*** 0.0176*** 0.0170*** 0.0195*** 0.0238*** 0.0202*** 0.0174*** 0.0207***
(8.22) (9.15) (10.08) (11.27) (8.11) (9.49) (9.76) (10.25)
Estimated Product Market Size (log transformed) 0.0128*** 0.0145*** 0.0123*** 0.0144*** 0.0126*** 0.0139*** 0.0121*** 0.0137***
(4.63) (5.66) (5.54) (4.48) (7.52) (8.31) (7.16) (6.39)
Firm Revenue (log transformed) 0.0215*** 0.0138*** 0.0106*** 0.00763*** 0.0208*** 0.0155*** 0.00852*** 0.00510***
(8.35) (8.27) (6.32) (3.59) (9.74) (8.65) (6.26) (4.33)
Planned Exit Channel Available? (Yes=1; No=0) 0.180*** 0.249*** 0.324*** 0.257*** 0.219*** 0.288*** 0.313*** 0.275***
(10.58) (11.23) (11.74) (10.43) (10.82) (10.33) (10.74) (11.53)
Pre-funding Startup Value Available? (Yes=1; No=0) 0.0627*** 0.0331** 0.0225** 0.00806 0.0825*** 0.0458*** 0.0205** 0.00837
(3.25) (2.34) (2.13) (1.44) (3.78) (3.19) (2.45) (1.28)
Difficult Level of Staffing 0.0193*** 0.0257*** 0.0242*** 0.0269*** 0.0175*** 0.0249*** 0.0260*** 0.0247***
(4.23) (3.55) (3.37) (4.25) (3.47) (4.22) (3.79) (4.13)
Average Number of Photos per Manager 0.0365*** 0.0284*** 0.0235*** 0.0266*** 0.0358*** 0.0281*** 0.0295*** 0.0274***
(3.73) (3.95) (4.26) (4.17) (3.68) (3.22) (3.10) (3.09)
U.S. Firm? (Yes=1; No=0) 0.0946*** 0.0835*** 0.0574*** 0.0455** 0.0929*** 0.0810*** 0.0573*** 0.0426**
(3.16) (3.55) (2.89) (2.33) (2.84) (3.05) (3.49) (2.36)
Startup Age -0.00543*** -0.00352*** -0.00180*** -0.00217*** -0.00539*** -0.00335*** -0.00216*** -0.00231***
(-3.24) (-2.93) (-3.17) (-3.28) (-3.27) (-3.19) (-3.25) (-3.04)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 0.129** 0.118*** 0.130*** 0.154** 0.103*** 0.173*** 0.116*** 0.158***
(2.18) (3.20) (3.39) (2.25) (3.37) (3.29) (3.16) (2.73)
Number of Observations 6870 6870 6870 6870 6870 6870 6870 6870
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.239 0.225 0.217 0.221 0.224 0.229 0.238 0.235
APPENDIX TABLE 4: General Analyses on Amount of Capital Raised (Robustness Check)
This table presents robustness checks on cross sectional regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. The dependent variable is the amount of capital raised from an equity crowdfunding campaign after log transformation. Due to log transformation, startups that receive no investment from crowdfunding are
excluded from analyses. Length of qualitative business introduction and percentage of promotional words used are sorted into 4 quantiles. Scale indicators from 1 to 4 in lieu of exact number or
percentage of words are used to measure the level of detail in qualitative business introduction and usage of promotional expression. Standard errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Predicted Percentage of Promotional Words Used in Scale
-1.230*** -1.255*** -1.264*** -1.249***
(-3.02) (-2.96) (-3.23) (-3.43)
Predicted Length of Qualitative Business Description in Scale 0.185*** 0.188*** 0.239*** 0.227***
(4.57) (5.69) (5.16) (5.43)
Number of Managers 0.166*** 0.139*** 0.145*** 0.147*** 0.168*** 0.159*** 0.142*** 0.138***
(4.62) (4.59) (4.18) (3.79) (5.66) (3.34) (2.71) (3.44)
Average Industry Experience 0.0323*** 0.0316*** 0.0279*** 0.0288*** 0.0294*** 0.0271*** 0.0280*** 0.0302***
(6.18) (5.73) (4.84) (4.69) (5.12) (4.34) (4.53) (4.14)
Average Educational Level 0.0645*** 0.0553*** 0.0584*** 0.0479*** 0.0562*** 0.0539*** 0.0575*** 0.0498***
(5.91) (5.58) (5.37) (4.28) (5.93) (5.16) (4.39) (3.85)
Planned Exit Channel Available? (Yes=1; No=0) 1.148*** 1.152*** 1.089*** 1.127*** 1.156*** 1.130*** 1.144*** 1.146***
(9.83) (9.72) (9.11) (8.86) (10.80) (9.66) (10.03) (9.87)
Estimated Startup Value (log transformed) 0.132*** 0.129*** 0.101*** 0.0928*** 0.135*** 0.112*** 0.0824*** 0.0902***
(4.29) (3.82) (4.03) (3.11) (3.85) (3.27) (3.09) (2.78)
Estimated Product Market Size (log transformed) 0.0501*** 0.0474*** 0.0496*** 0.0422*** 0.0483*** 0.0419*** 0.0426*** 0.0407***
(7.96) (7.25) (8.34) (7.40) (8.58) (8.23) (7.75) (7.34)
Firm Revenue (log transformed)
0.0894*** 0.0913** 0.0946***
0.0867*** 0.0835*** 0.0821***
(8.60) (8.39) (7.74)
(9.07) (8.25) (8.32)
Percentage of R&D Expense
0.278** 0.315** 0.291***
0.282*** 0.256*** 0.235**
(2.46) (2.33) (2.85)
(2.60) (2.73) (2.15)
U.S. Firm? (Yes=1; No=0)
0.0154 0.0149
0.0753 0.0756
(0.83) (0.67)
(1.31) (1.08)
Startup Age
0.0317*** 0.0365***
0.0382*** 0.0396***
(3.41) (3.57)
(3.29) (3.40)
Difficult Level of Staffing
-0.00258
-0.00306*
(-1.06)
(-1.69)
Average Number of Photos per Manager
0.0564
0.0687
(1.29)
(1.23)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? Yes Yes Yes Yes Yes Yes Yes Yes
Constant 1.122*** 0.986*** 0.853*** 1.094*** 1.108*** 1.036*** 1.074*** 1.062***
(4.37) (4.06) (3.58) (3.60) (3.65) (3.97) (4.05) (3.43)
Number of Observations 6594 6594 6594 6594 6594 6594 6594 6594
Number of Clusters (industry classification) 20 20 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115 115 115
R2 0.184 0.198 0.215 0.216 0.169 0.187 0.204 0.213
APPENDIX TABLE 5: Subsample Analyses on Amount of Capital Raised (Robustness Check)
This table presents robustness checks on cross sectional regressions evaluating the impact of qualitative business descriptions and usage of promotional words on equity crowdfunding outcome. Only fully funded equity crowdfunding projects are included in the analyses. The dependent variable is the amount of capital raised from an equity crowdfunding campaign after log transformation. Length of
qualitative business introduction and percentage of promotional words used are sorted into 4 quantiles. Scale indicators from 1 to 4 in lieu of exact number or percentage of words are used to measure
the level of detail in qualitative business introduction and usage of promotional expression. Standard errors are clustered by startup industry and crowdfunding campaign initiation month. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Predicted Percentage of Promotional Words Used in Scale
-1.103*** -1.025** -1.047**
(-2.76) (-2.45) (-2.38)
Predicted Length of Qualitative Business Description in Scale 0.422*** 0.373*** 0.329***
(12.76) (13.40) (12.19)
Number of Managers 0.168*** 0.172*** 0.153*** 0.216*** 0.185*** 0.164***
(5.15) (5.26) (5.21) (4.72) (4.59) (4.11)
Average Industry Experience 0.0645*** 0.0472** 0.0440*** 0.0522*** 0.0429*** 0.0381**
(4.05) (2.46) (2.88) (2.81) (2.80) (2.49)
Average Educational Level 0.0483 0.0355 0.0221 0.0626 0.0537 0.0408
(1.42) (0.89) (0.48) (0.95) (0.76) (0.83)
Estimated Startup Value (log transformed) 0.183*** 0.174*** 0.177*** 0.169*** 0.177*** 0.165***
(6.42) (6.13) (5.39) (7.10) (6.34) (6.27)
Estimated Product Market Size (log transformed) 0.135*** 0.128*** 0.125*** 0.141*** 0.136*** 0.134***
(11.65) (12.14) (12.37) (10.93) (9.85) (10.08)
Firm Revenue (log transformed) 0.162*** 0.153*** 0.148*** 0.180*** 0.185*** 0.169***
(3.28) (2.79) (2.63) (2.74) (2.85) (2.68)
Percentage of R&D Expense
-0.329 -0.268
-0.230 -0.224
(-1.45) (-1.17)
(-1.09) (-0.78)
Startup Age
0.00469
0.00583
(0.74)
(0.91)
Control for Equity Retention Ratio? Yes Yes Yes Yes Yes Yes
Control for Video Used in Fundraising Campaign? Yes Yes Yes Yes Yes Yes
Control for Amount of Capital Seeking? No No No No No No
Constant 1.083*** 1.107*** 0.976*** 1.694*** 1.711*** 1.709***
(3.89) (3.85) (3.66) (3.91) (3.26) (3.25)
Number of Observations 2497 2497 2497 2497 2497 2497
Number of Clusters (industry classification) 20 20 20 20 20 20
Number of Clusters (account creation month) 115 115 115 115 115 115
R2 0.142 0.143 0.144 0.131 0.131 0.132
APPENDIX TABLE 6: Effect of Qualitative Information on Matched Sample
This table presents estimated effect of qualitative business information on fundraising outcome, measured by percentage of fundraising plan completed (Panel A) and amount of capital raised from an equity crowdfunding campaign (Panel B). Observations are first divided into two groups based on whether the length of qualitative startup information is above group median. The group with more
qualitative information is denoted as treatment group; less qualitative information, control group. Observations in the treatment group are then matched with observations in the control group based on
control variables specified in Section V. The matching process is based on propensity score matching (PSM), nearest-neighbor matching (NNM) or inverse probability weighted regression adjustment (IPWRA) method. Mahalanobis distance is used in NNM. Logistic treatment models are used in PSM and IPWRA. *, **, *** indicate significance at the 10%, 5%, and 1% levels, respectively.
Panel A. Percentage of Fundraising Plan Completed
Model 1 Model 2 Model 3
Coefficient 0.0552*** 0.0425*** 0.0460***
Robust Standard Error 0.00958 0.00542 0.00766
Z Value 5.76 7.83 6
P Value 0 0 0
Estimator Propensity Score Matching Nearest-neighbor Matching Inverse Probability Weighted Regression
Adjustment
Estimate Average Treatment Effect On the Treated On the Treated On the Treated
Number of Observations 6870 6870 6870
Panel B. Amount of Money Raised
Model 4 Model 5 Model 6
Coefficient 0.226*** 0.262*** 0.241***
Robust Standard Error 0.0375 0.0493 0.0531
Z Value 6.03 5.31 4.54
P Value 0 0 0
Estimator Propensity Score Matching Nearest-neighbor Matching Inverse Probability Weighted Regression
Adjustment
Estimate Average Treatment Effect On the Treated On the Treated On the Treated
Number of Observations 6870 6870 6870