Quality Revealing or Overstating? -...

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

[email protected]

Yelin Zhang

School of Business Administration - Gonzaga University

502 East Boone Avenue

Spokane, WA 99258-0102

[email protected]

* 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.

mailto:[email protected]

mailto:[email protected]

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, 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, 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

https://papers.ssrn.com/sol3/Papers.cfm?abstract_id=2447567

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2882026

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 0.80)


-0.425** -0.736*** -1.023*** -1.158***

(-2.16) (-4.65) (-3.42) (-4.14)


(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)


(3.16) (4.02) (2.94) (3.65) (2.66) (4.27) (4.36) (3.72)


(7.16) (9.38) (9.59) (11.64) (8.28) (9.46) (10.73) (11.28)


(4.68) (5.96) (5.75) (5.15) (6.75) (7.24) (6.43) (5.94)


(8.26) (8.15) (6.17) (4.02) (8.73) (8.28) (6.29) (3.97)


(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)




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)



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.



-0.937** -1.274*** -1.262*** -1.245***

(-2.49) (-2.80) (-3.26) (-3.25)


(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)


(6.33) (5.45) (5.03) (4.64) (5.93) (4.57) (4.67) (3.92)


(5.16) (5.25) (4.37) (3.84) (4.92) (5.09) (4.26) (3.50)


(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)


(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)


0.323*** 0.306** 0.287**

0.274*** 0.266** 0.249**

(2.72) (2.33) (2.15)

(2.63) (2.39) (2.04)


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)


-0.00246

-0.00324

(-1.44)

(-1.51)


0.0678

0.0736

(0.43)

(0.31)




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)




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


-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)


-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.


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)


(5.06) (4.37) (4.28) (5.29) (4.35) (4.29) (4.16) (4.24)


(7.64) (7.15) (8.03) (9.81) (10.22) (11.46) (10.28) (9.62)


(5.38) (5.46) (5.25) (4.42) (7.63) (6.85) (7.14) (6.13)


(11.17) (10.03) (10.65) (10.24) (9.34) (9.57) (8.85) (8.69)


(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)


0.0414*** 0.0326*** 0.0297***

0.0242*** 0.0237*** 0.0236***

(3.74) (4.63) (4.15)

(4.60) (3.63) (4.25)


0.0815*** 0.0784***

0.0875*** 0.0760***

(3.22) (3.35)

(2.76) (2.81)


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)


0.0264

0.0302

(1.13)

(1.20)




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)




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)


-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)


(3.28) (3.55) (4.04) (3.78) (3.44) (3.45) (3.28) (4.13)


(8.22) (9.15) (10.08) (11.27) (8.11) (9.49) (9.76) (10.25)


(4.63) (5.66) (5.54) (4.48) (7.52) (8.31) (7.16) (6.39)


(8.35) (8.27) (6.32) (3.59) (9.74) (8.65) (6.26) (4.33)


(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)




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)



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.



-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)


(6.18) (5.73) (4.84) (4.69) (5.12) (4.34) (4.53) (4.14)


(5.91) (5.58) (5.37) (4.28) (5.93) (5.16) (4.39) (3.85)


(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)


(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)


0.278** 0.315** 0.291***

0.282*** 0.256*** 0.235**

(2.46) (2.33) (2.85)

(2.60) (2.73) (2.15)


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)


-0.00258

-0.00306*

(-1.06)

(-1.69)


0.0564

0.0687

(1.29)

(1.23)




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)




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


-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)


-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

Quality Revealing or Overstating? -...

Documents

Transcript of Quality Revealing or Overstating? -...