Startup Financing with Patent Signaling under Ambiguity

• Date: 01 February 2017
• DOI: http://doi.org/10.1111/ajfs.12162
• Published date: 01 February 2017

Startup Financing with Patent Signaling

under Ambiguity*

Guangsug Hahn

Division of Humanities and Social Sciences, POSTECH, Republic of Korea

Kwanho Kim

Department of Economics, Chungbuk National University, Republic of Korea

Joon Yeop Kwon**

Division of Humanities and Social Sciences, POSTECH Republic of Korea

Received 31 August 2016; Accepted 12 December 2016

Abstract

One of the most important challenge s for startup companies is securing financing. Indeed,

it is crucial for startups to demonstrate the ir projects’ profitability to potential inv estors.

We develop a model of single-stage st artup financing with signaling unde r ambiguity. Nat-

ure determines the ability of a tec hnology entrepreneur (startup), wh o strategically chooses

a costly patent level to signal hi s ability to potential investors. Because the project under-

taken by a startup may involve highly inn ovative technology and may not be well know n

to agents, they might face ambiguity ab out the value of the project. To examine ambigui ty

effects on startup financing, we provide three di fferent financing models in view of the

degree of ambiguity: (i) no ambigui ty; (ii) only investors face ambigui ty; (iii) all agents

face ambiguity. In each model, we de rive perfect Bayesian equilibria an d refine them into

a unique equilibrium by imposing the Intui tive Criterion of Cho and Kreps (Qu arterly

Journal of Economics, 102, 1987, 1 79) or its extension. We analyze the re fined equilibria

from the perspectives of agents’ equity sha res and expected profits, and equilibriu m patent

levels.

Keywords Startup financing; Signaling; Perfect Bayesian equilibrium; Intuitive Criterion

JEL Classification: D82, G14, G24

*This paper received the Asia-Pacific Journal of Financial Studies Best Paper Award at the

11th International Conference on Asia-Pacific Financial Markets. We would like to thank Jae

Joon Han, Sooyoung Song, and Eunjung Yeo for valuable comments, which have substan-

tially improved our paper. This work (Grants No. 10031496) was supported by the Business

for University Entrepreneurship Center funded by the Korea Small and Medium Business

Administration in 2015.

**Corresponding author: Joon Yeop Kwon, Division of Humanities and Social Sciences,

POSTECH, Korea. Tel: +82-54-279-2729, Fax: +82-54-279-3699, email: ykwon@postech.ac.kr.

Asia-Pacific Journal of Financial Studies (2017) 46, 32–63 doi:10.1111/ajfs.12162

32 ©2017 Korean Securities Association

1. Introduction

One of the most important challenges for startup companies is securing financing.

Owing to the absence of track records, it is crucial for startups to signal their pro-

jects’ profitability to potential investors. In other words, startups need to reveal reli-

able information about their ability to attract investors during the early financing

stages. For technology startups, the number of filed patents can be a useful signal to

access seed investors. As Graham et al. (2009) point out, technology startups tend

to hold patents for a competitive advantage, to secure financing, and enhance their

reputation. In particular, by analyzing the Berkely Patent Survey, they find that it is

easier for startups to attract funding from external investors if they hold more

patents. Conti et al. (2013a) empirically show that, in startup financing, an increase

in the level of patents raises both the frequency and number of investments from

venture capitals. They explain this empirical observation by using a signaling game

in which players consider a two-dimensional signal that consists of patent level and

investments from acquaintances. Conti et al. (2013b) employ a single-stage financ-

ing model in which the entrepreneur uses patent signals to signal his ability and

empirically show that a startup’s patents level is endogenously determined.

1

Another important issue for startups concerns ambiguity about project value,

which is the uncertainty regarding the true success probability of a startup’s project.

For instance, outside investors may not easily obtain exact information about the

entrepreneur’s true success probability, and therefore regard it as a random variable.

In this case, we say that investors face ambiguity about the entrepreneur’s project

value. Often, a startup’s project may be innovative and has sparse track records, so

agents (i.e., the entrepreneur and investors) must make decisions without sufficient

information about the entrepreneur’s ability. In a different context, Rigotti et al.

(2008) point out that technology startups often have ambiguous information about

their own project value. However, to the best of our knowledge, no existing model

of statup financing considers ambiguity effects.

It is practically important to examine the effects of ambiguity when a startup

uses its number of patents to signal its success probability to investors. One can ask

the following questions. Compared with cases in which agents are well-informed

about a project, under ambiguity: How do they make decisions? How does the

entrepreneur signal his ability to potential investors? How do investors require their

compensation and how are equity shares distributed? One may conjecture that an

entrepreneur acquires more patents with the goal of demonstrating the profitability

of his project and investors require more compensation under ambiguity than

otherwise. However, the validity of the conjecture depends on who faces ambiguity

1

Elitzur and Gavious (2003) and Kim and Wagman (2016) consider different kinds of signal-

ing devices in two-stage startup financing models. In Elitzur and Gavious (2003), whether an

entrepreneur approaches an angel or not is a signal about his future effort level. In Kim and

Wagman (2016), the entrepreneur’s decision about whom he makes a contract with in the

first stage is a signal to investors in the second stage.

Startup Financing with Patent Signaling under Ambiguity

©2017 Korean Securities Association 33

and what the levels of the underlying parameters are (see Proposition 5.2). The pur-

pose of this paper is to analyze how a startup’s patent signaling affects early-stage

financing under ambiguity. To do this, we provide single-stage startup financing

models in which an entrepreneur strategically chooses a costly patent level as a sig-

naling device to inform his ability or success probability to potential investors.

Investors participate in seed investment to initiate the entrepreneur’s pro ject after

observing the entrepreneur’s patent level. Similar to Spence (1973), for simplicity,

we assume that the entrepreneur’s project value is not affected by the number of

patents held.

To analyze the decision making under ambiguity, we employ the smooth ambi-

guity model of Klibanoff et al. (2005), which represents preferences by the expected

distortion of the expected utility; we consider its special case in which agents are

risk-neutral and ambiguity-neutral.

2

To examine the effects of ambiguity, we pro-

vide three models that differ in degree of ambiguity: (i) no ambiguity; (ii) only

investors face ambiguity; (iii) all agents face ambiguity. In the first model (bench-

mark model), the entrepreneur exactly knows his own type or probability of true

success, which investors cannot observe. This model represents the case in which

the project involves a well-known technology. In the second model (Model I), the

entrepreneur also exactly knows his own type. However, investors face ambiguity

about project value, that is, they know only the intervals that can contain the entre-

preneur types, but they cannot observe which interval contains the entrepreneur’s

type. This model supports the case in which the project involves an intermediate-

level innovation. In the third model (Model II), even the entrepreneur does not

exactly know his own type. Indeed, when the project involves a high-level innova -

tion, the entrepreneur cannot know his own probability of success. Here, both the

entrepreneur and investors face the same ambiguity as investors do in Model I, but

the entrepreneur recognizes the interval that contains his own type.

We derive perfect Bayesian equilibria (PBE) in the signaling game of each star-

tup financing model and refine them into a unique equilibrium by imposing the

Intuitive Criterion of Cho and Kreps (1987) or its extension. Then we analyze the

refined equilibria from the perspectives of agents’ equity shares, patent level

acquired by the entrepreneur, and his expected profit. It is noted that, since we

assume Bertrand competition in the investment market, the investor’s profit is zero

in each refined equilibrium. We find that the entrepreneur should acquire the great-

est number of patents to signal his ability to investors when only investors face

ambiguity (i.e., in Model I). This result occurs because investors more conserva-

tively respond to the entrepreneur’s signaling than when all agents resolve or face

it. However, investors can be reasonably expected to demand a larger equity share

when the project is not well known to them than otherwise. However, we find that

2

One can employ alternative ambiguity models such as the maximin expected utility model

(Gilboa and Schmeidler, 1989), the multiplier model (Hansen and Sargent, 2001), and the

variational preference model (Maccheroni et al., 2006).

G. Hahn et al.

34 ©2017 Korean Securities Association