Trade secret laws and initial public offering underpricing

Abstract

This paper examines the impact of trade secret laws on the underpricing of initial public offerings (IPOs) in the United States, where we focus on the Uniform Trade Secret Act and/or the inevitable disclosure doctrine. Given that trade secret laws help firms protect their know-how in the form of trade secrets, we propose that trade secrets protected by these laws may result in increased corporate opacity, leading to greater IPO underpricing. Empirically, the average first-day returns of IPOs in states with trade secret laws are 12.7% higher than those in states without such laws. Using the information disclosed in the 10-K report on whether the firm owns trade secrets, we suggest that the existence of trade secret laws enhances the protection effect of firm maintenance of trade secrets, which leads to greater IPO underpricing. Further analyses show that the effect of trade secret laws is reinforced for IPO firms operating in complex industries, IPO firms with R&D investments, and large IPO firms, suggesting that trade secret laws increase the difficulties investors face in assessing the intrinsic value of such firms.

1 Introduction

Corporate innovation plays an important role in initial public offerings (IPOs), given the prevalence of technology firms in the U.S. IPO market. According to Bernstein (2015), about 40% of U.S. IPO firms since the 1980s have been technological firms. Scholars began to pay more attention to the relationship between innovation and IPOs after the internet bubble in late 1990s. Loughran and Ritter (2004) find that the average first-day return of the IPOs (most of them are technology firms) during the internet bubble period is 65%, significantly higher than first-day returns in other periods. Other papers also examine the role of corporate innovations on IPO underpricing and post-IPO performance, and find that more R&D, which is strongly related to the degree of information asymmetry, results in greater IPO underpricing (e.g., Boone et al. 2016; Zhou and Sadeghi 2019). However, some scholars suggest that patent-based innovation and technology may attract more favorable media coverage and reduce IPO underpricing in an environment with transparent information (Heeley et al. 2007; Bhattacharya et al. 2009).¹

In this paper, we investigate the role of trade secret laws in IPO underpricing. Though previous studies have explored the relationship between innovation and the IPO performance, most of them focus on R&D expenditure and patent-based innovations. A legally precise and broadly defined term of innovation is intellectual property rights. The World Intellectual Property Organization (WIPO) states that intellectual property rights refer to “creations of the mind, such as inventions; literary and artistic works; designs; and symbols, names and images used in commerce”.² Intellectual property rights encompass a range of legal protections, such as patents, trademarks, trade secrets, and copyrights. Among these, trade secrets are regarded as the most ancient form of intellectual property rights. Examples of trade secrets include confidential formulas, production processes, customer lists, and other undisclosed details critical to the functioning of the entity. According to the report of WIPO, 52.3% of innovative firms use trade secrets to protect their know-how while 31.7% of innovative firms use patent systems to protect intellectual property rights in 24 European Union countries.³ Thus, our paper addresses a gap in the literature by improving our understanding of the effect of intellectual property rights on IPO underpricing.

In this paper, we propose that trade secret laws enhance trade secret legal protections, increasing corporate opacity, which in turn leads to greater IPO underpricing. Trade secrets are not filed in the patent system. Instead, trade secrets represent non-disclosed know-how and technologies and reduced knowledge flow from innovative firms to outsiders (Wang 2023). Since the trade secrets of a firm are usually unobservable, previous studies (e.g., Glaeser 2018; Klasa et al. 2018; Li et al. 2018) rely on trade secret legal protections to determine to what extent the firm is willing/able to generate trade secrets. Given the assumption that trade secret laws encourage firms to protect their innovations with trade secrets, the degree of information asymmetry would increase for firms located in states with trade secret laws. By the same token, the IPO literature (e.g., Ritter and Welch 2002; Ljungqvist 2007) suggests that IPO underpricing is more severe when the degree of information asymmetry of the IPO firm is higher for several reasons, including avoiding winner’s curse, informational cascade, and book-building theory. Therefore, we hypothesize that IPO firms experience greater IPO underpricing when they are located in states with trade secret laws.

We collect U.S. IPO data from Securities Data Company (SDC) Platinum. Accounting and stock market data are obtained from Compustat and The Center for Research in Security Prices (CRSP) databases. Firms in the U.S. follow two sets of trade secrets laws at the state level: the Uniform Trade Secret Act (UTSA) and Inevitable Disclosure Doctrine (IDD). Information about trade secrets laws is available in Png (2017a, 2017b), Glaeser (2018), Klasa et al. (2018), and Li et al. (2018). We adopt the UTSA and IDD in this paper because the year when a state passed the UTSA and/or IDD varies across states, allowing a time-series and cross-sectional heterogeneity design that better captures the effect of trade secret laws on IPO underpricing.

Measuring the IPO underpricing by the first-day return at the IPO day, we find that the average first-day return is 14.3% for IPO firms located in states without trade secret laws. By contrast, average first-day returns are ranged from 19.8% to 26.6% for IPO firms located in states with the UTSA and/or IDD. We also perform a regression analysis, which shows that the average first-day return of IPOs in states with trade secret laws is 12.7% higher than those in states without such laws, using a battery of control variables and industry and state fixed effects. The return difference of 12.7% explains about 52.4% of average first-day returns of IPOs in the U.S. We perform robustness checks to prevent domination of our results by the influence of specific firms. After excluding California-based firms, penny stocks, small firms, and controlling for location choice issues, we still obtain consistent results. We use the disclosure information of whether the firm has trade secrets in the 10-K report for further analysis because trade secret law is effective in providing protection only when a firm has trade secrets.⁴ We once again confirm our hypothesis and find that IPO firms with disclosure that they own trade secrets earn higher first-day returns.

Next, we study the moderating effects of trade secret laws on IPO underpricing by examining covenant not to compete (CNC) agreements, complex industries, patents, R&D intensity, and firm size. First, trade secret laws and CNC agreements may substitute for each other in protecting an employer’s innovation (Png 2017a). Second, industries with complex products and technologies may hold more valuable technologies as trade secrets, reducing the likelihood of reverse-engineering their technologies (Cohen et al. 2000; Png 2017a). Third, we examine the role of patent systems because patents are another important legal system that protect a firm’s intellectual property. Fourth, we investigate whether compared with non-R&D firms with trade secrets, R&D firms with trade secrets behave differently in their IPO offering procedures. Fifth, we test the role of firm size, which may affect the degree of information asymmetry. Our empirical results show that trade secret laws have a more pronounced impact on IPO underpricing for firms in more complex industries, firms with R&D expenses, and large IPO firms.

This study contributes to the literature in two ways. First, recent papers have examined the relationship between IPOs and corporate innovations, particularly focusing on R&D and patent-based innovations (e.g., Heeley et al. 2007; Bhattacharya et al. 2009; Guo and Zhou 2016; Boone et al. 2016; Zhou and Sadeghi 2019). More importantly, the literature does not offer conclusive findings about how innovation may affect IPO underpricing. For example, Chin et al. (2006) uncover a positive relationship between the number of patents of an IPO firm and corresponding IPO underpricing. Yet, Heeley et al. (2007) find that patent-based innovation may reduce IPO underpricing in an environment with transparent information but induce IPO underpricing in an environment with opaque information. Yang and Yuan (2022) suggest that trademarks of the IPO firm can reduce IPO underpricing. Our paper explores the research question of how intellectual property rights in the form of trade secrets may affect IPO underpricing. Thus, our paper addresses an unexplored gap in the IPO literature on trade secrets.

Second, our study complements the trade secret literature by exploring how it could relate to IPO underpricing. We clearly understand the impact of trade secret legal protections on firm decisions, such as capital structure, disclosure, and acquisitions. Trade secret laws can alter firms’ decision to voluntarily disclose proprietary and nonproprietary information, increasing corporate leverage and the likelihood of being an acquisition target (e.g., Glaeser 2018; Klasa et al. 2018; Li et al. 2018; Chen et al. 2020a). Our study provides empirical support for the contention that trade secret laws can affect the pricing decision in the IPO process.

The remaining sections are as follows. Section 2 describes trade secret laws in the U.S. and presents the literature review and hypothesis development. Section 3 presents the data, variables, and methodology. Section 4 offers the empirical results. This paper concludes with Sect. 5.

2 Literature reviews and hypothesis development

2.1 Trade secrets laws in the U.S. and related studies

The World Intellectual Property Organization (WIPO) defines trade secrets as “intellectual property rights on confidential information that may be sold or licensed.” In order to be classified as a trade secret, certain conditions must be met. First, the knowledge in question must possess commercial value by virtue of its secrecy. Second, it should be known exclusively to a limited group of individuals. Last, the rightful holder of the trade secret must have implemented reasonable measures to maintain its confidentiality, which may include the utilization of confidentiality agreements with business partners and employees.⁵ Moreover, unlike patents and trademarks, companies are not required to file trade secrets with any government institute (e.g., United States Patent and Trademark Office). Therefore, trade secret legal protection involves safeguarding undisclosed and confidential know-how and technology.

Even if a firm does not file trade secrets with a government institute, corporate trade secrets are still protected by trade secret laws in the U.S. These laws aim to prevent the misuse and infringement of a firm's trade secrets. In cases of trade secret misappropriation, courts have the authority to order remedies and injunctions. Before granting such relief, courts may encourage settlements between the defendant and the plaintiff involved in a trade secret litigation. For instance, in 2020, Uber and Waymo reached a settlement of $245 million to resolve a trade secret dispute. While legal protection for trade secrets increases the cost of theft and discourages the publication of sensitive information regarding crucial technologies (Hall et al. 2014), it can also hinder labor mobility and the expansion of technology. Consequently, there are numerous debates about the necessary extent of government intervention in trade secret legal protection (Wang 2023).

Trade secret legal protections in the U.S. began to evolve in the late 1970s. As trade secrets are governed by common law, the level of legal protection afforded to a company’s trade secrets depends on court orders and decisions, particularly those made by U.S. federal District Courts and Courts of Appeals. One significant doctrine for the court’s judgment is the Inevitable Disclosure Doctrine (IDD), which states that departing staff and management in certain states (e.g., Massachusetts and New York) will inevitably disclose all trade secrets they have witnessed during their employment. To prevent the inevitable disclosure of trade secrets to competitors by departing personnel and management, a competitor that recruits persons with knowledge of trade secrets may violate the IDD and infringe on trade secrets without proving actual theft (for instance, by photocopying the trade secret documents).

Moreover, UTSA was authorized in 1979 by the National Conference of Commissioners on Uniform State Laws. The UTSA, created by the Uniform Law Commission in 1979 and revised in 1985, aims to establish uniform laws across the United States regarding trade secrets. Since the 2000s, a majority of U.S. states have adopted UTSA, ensuring a consistent legal framework for trade secrets. This standardization is particularly beneficial for companies operating in multiple states, simplifying the legal landscape. The UTSA is intended to protect trade secrets from being appropriated by improper means, where the term “improper” indicate not only illegal activities but also other improper lawful conducts under the circumstances (Wang 2023). However, the UTSA states that improper means does not include reverse engineering (Glaeser 2018). Practically, potential infringed firms could accuse other firms of trade secrets theft in a state court with jurisdiction, which is usually related to whether or not the plaintiffs operate and/or conduct commercial activities in the state (Almeling et al. 2010; Effron 2016). In the U.S., a dual-track system allows trade secret owners to pursue legal action for trade secret infringement through UTSA and/or IDD.

Table 1 displays the enactment years of the UTSA and IDD in each state (including the Washington, D.C.) up until 2020. Before 2000, most states implemented UTSA, and only 7 states did not adopt UTSA. As of 2020, 20 states have implemented both the UTSA and IDD. UTSA has been adopted by 49 states and Washington, D.C., with the exception New York. New York is the only state to adopt the IDD but do not to adopt the UTSA. The implementation of UTSA and/or IDD in all U.S. states by 2020 demonstrates each state’s commitment to passing laws to protect the trade secrets of corporate innovation. This heterogeneity in trade secret legal protection across states and over time allows us to investigate the impact of trade secret laws on IPO underpricing.

Table 1 Enactment and adoption of UTSA and IDD by state (up to 2020)

In 2016, the United States passed the Defend Trade Secrets Act (DTSA) as a federal law, which involves both civil and criminal actions in federal court for the misappropriation of trade secrets. As its criminal law content, the DTSA is enacted as amendments to a criminal law of the Economic Espionage Act. As for its civil law content, the DTSA emphasizes Congress’ alignment with UTSA, which has been adopted in some form by nearly every state in the U.S. Accordingly, the DTSA creates a uniform standard for trade secrets across states and provides a federal civil remedy for trade secret misappropriation. DTSA provides additional protection and remedies for trade secret owners, including the ability to file lawsuits in federal courts. Finally, the UTSA is not preempted by the DTSA. Instead, plaintiffs can sue for trade secret infringements in state court or federal court, or both.

Recent papers study the impact of trade secret laws on corporate decisions (e.g., Png 2017a, b; Glaeser 2018; Klasa et al. 2018; Li et al. 2018; Wang 2023). Png (2017a) finds that stronger trade secrets laws are associated with greater innovation, highlighting the role of legal protection in promoting innovation. Png (2017b) reveals a trade-off between secrecy and patenting strategies, suggesting that stronger trade secrets protection leads to decreased patenting activity. Glaeser (2018) shows that firms with valuable proprietary information publicly disclose less information, indicating that trade secrets protection may hinder corporate transparency. Klasa et al. (2018) find that trade secrets laws are linked to higher leverage ratios and decreased external financing, implying that firms with effective trade secrets protection rely more on internal resources for financing. Li et al. (2018) provide evidence indicating that firms facing higher proprietary costs disclose less information, supporting the proprietary cost hypothesis. This suggests that strengthening the protection of trade secrets may limit the disclosure practices. Wang (2023) demonstrates that trade secrets laws are associated with reduced technology spillovers, suggesting that firms may be less willing to share knowledge due to increased legal protection. These findings contribute to understanding the dynamics of legal protection, disclosure practices, and innovation outcomes in trade secrets contexts.

2.2 Innovation and IPO underpricing

Underpricing of IPOs occurs when an IPO’s offering price is lower than its first-day closing price, indicating money left on the table to the underwriter and investors (Loughran and Ritter 2004). IPO underpricing is often attributed to existence of information asymmetry, in which external investors lack adequate information about primary market companies to determine the true value of shares. This perception of increased risk due to a lack of information could lead investors to demand a price reduction for the IPO as compensation. Detailed theories could be found in the survey paper of Ritter and Welch (2002).

Furthermore, the success of innovation is contingent on the ability of firm managers to transform new technologies into profitable products and services. However, increasing R&D expenditures does not inherently result in successful innovations. Typically, intellectual property rights such as patents, trademarks, trade secrets, and copyrights are used to demonstrate R&D strengths, but investors find it challenging to assess the intrinsic value of a company.⁶ Research indicates that an increase in R&D correlates considerably with the degree of information asymmetry and leads to greater underpricing of IPOs. Boone et al. (2016) and Zhou and Sadeghi (2019) confirm this argument and find greater IPO underpricing for firms with more R&D expenditures. By contrast, Heeley et al. (2007) and Bhattacharya et al. (2009) argue that patent-based innovation and technology can attract more favorable media coverage and reduce IPO underpricing. Yang and Yuan (2022) find that IPO underpricing is greater when the IPO firm registers more trademark, which could be a proxy for product innovations (Chen et al. 2022a). Therefore, whether the innovation of a firm is positively related to first-day IPO returns remains inconclusive.⁷

2.3 Hypotheses development

The use of trade secrets can lead to a lack of transparency and limited information disclosure, making it challenging for investors to accurately assess the intrinsic value of the company and the potential risks associated with investing in it. This causes the information asymmetry and may lead to IPO underpricing. Because trade secrets are frequently unobservable, previous research (e.g., Glaeser 2018; Klasa et al. 2018; Li et al. 2018) has relied on trade secret legal protections and investigate to what extent the trade secrets may affect corporate decision. Given the assumption that trade secret laws provide protection for firms that safeguard their know-how through trade secrets, the presence of such laws would lead to an increase in information asymmetry for firms located in states with trade secret laws. By the same token, the IPO literature (e.g., Ritter and Welch 2002; Ljungqvist 2007) suggests that the degree of IPO underpricing increases with the level of information asymmetry, explained by theories such as avoiding winner’s curse, informational cascade, and book-building theory. Therefore, given that trade secret law is beneficial for firms to protect their know-how in the form of trade secrets, we hypothesize that when IPO firms are located in states with trade secret laws, they experience greater IPO underpricing due to greater information asymmetry.

Hypothesis 1

IPO underpricing is more pronounced in states with trade secret laws than in states without trade secret laws.

The second hypothesis discusses the role of covenant not to compete (CNC) agreements on the relationship between trade secret laws and IPO underpricing. A CNC serves as a legal contract that imposes limitations on ex-employees, preventing them from engaging in employment with a rival within a defined time frame or geographical range. This strategic measure aids businesses in safeguarding their proprietary information (Gilson 1999). By the same token, trade secrets can also prevent departing employees from bringing the know-how of ex-employer to rivals (Wang 2023). Hence, trade secret laws and CNC agreements may be substitutes as employers choose to sue for violations of use of confidential information. In this paper, we leverage the index of CNC in Table A1 of Garmaise’s (2011) CNC index and partition the IPO firms into firms located in states with high and low CNC indexes. The effect of the trade secret laws should be stronger for IPO firms located in state with low CNC indexes. We thus hypothesize that the relationship between trade secret laws and IPO underpricing will be weakened when the firm is located a state with CNC laws.

Hypothesis 2

The effect of trade secret laws on the IPO underpricing is more pronounced when the IPO firm is located in a state with weak CNC laws.

In addition, we posit that the intricacies of an industry can amplify the impact of trade secret laws on the IPO underpricing. In industries characterized by discrete and traditional practices, even if a firm protects trade secrets, competitors may gain access to the same knowledge through reverse engineering. By contrast, in technologically sophisticated industries, outsiders face increasing challenges in comprehending complex technological nuances. Building on this perspective, Png (2017a) suggests and finds that industries marked by complexity, such as telecommunications equipment or semiconductors, exhibit enhanced efficacy of trade secret laws. Therefore, we propose that IPO firms operating in more complex industries may exhibit greater underpricing compared to their peers in less complex industries, particularly when these IPO firms benefit from the protection of trade secret laws.

Hypothesis 3

The effect of trade secret laws on the IPO underpricing is more pronounced when the IPO firm operates in a more complex industry.

Moreover, we conjecture that patent systems may influence the effect of trade secret laws on IPO underpricing. Two predictions exist regarding this effect. On one hand, patents and trade secrets act as substitutes within a given technology. Specifically, a company cannot simultaneously patent a technology and designate it as a trade secret, given the latter’s confidentiality requirements. From this standpoint, an IPO firm heavily reliant on the patent system should downplay the significance of trade secrets, thereby weakening the impact of trade secret laws on IPO underpricing. On the other hand, companies typically file patents for certain technologies while keeping others as trade secrets. For instance, in the case of process innovation, the efficacy of intellectual property right protection is higher under trade secret laws than patent laws. Therefore, from a firm-level perspective, patents and the use of trade secrets are complementary. Given this, we propose the hypotheses outlined below.

Hypothesis 4a

The effect of trade secret laws on the IPO underpricing is more pronounced when the IPO firm files more patents.

Hypothesis 4b

The effect of trade secret laws on the IPO underpricing is more pronounced when the IPO firm files fewer patents.

Furthermore, R&D investment can play an important role in the relationship between IPO underpricing and trade secret laws. Firms hold a preponderance of trade secrets, including intricate technological details, process innovations, formulas, and algorithms—all intricately linked to their R&D investments (Varadarajan 2014). Given that R&D is often viewed as an intangible component that contributes to firm valuation, trade secrets may have greater value relevance when the IPO firm is actively involved in R&D activities. Therefore, R&D investment may strengthen the impact of trade secret laws on IPO underpricing. We propose the corresponding hypotheses as follows.

Hypothesis 5

The effect of trade secret laws on the IPO underpricing is more pronounced for IPO firms with R&D spending.

Finally, we evaluate the impact of firm size on the relation between IPO underpricing and trade secret legal protection. In general terms, the size of a firm affects the degree of information asymmetry displayed by IPOs. Smaller firms, characterized by heightened information asymmetry, are prone to encountering pronounced IPO underpricing, regardless of whether these smaller IPOs possess trade secrets. Accordingly, the additional impact of trade secret laws is anticipated to be less substantial for smaller IPOs. Larger firms often receive attention from the public, media or analysts, and the information asymmetry problem between investors and firms is less serious. Therefore, the additional information asymmetry problem caused by a trade secret law that prohibits the disclosure of commercial confidential information may be more substantial for large firms with more transparent information than for smaller firms. We propose the following hypothesis suggesting an enhanced incremental effect of trade secret laws on large IPOs.

Hypothesis 6

The effect of trade secret laws on the IPO underpricing is more pronounced for large IPO firms.

3 Data and summary statistics

3.1 Data

We collect U.S. IPO firms from the Securities Data Company (SDC) Platinum from 1994 to 2020. We require that these IPO firms are covered in the Center for Research in Security Prices (CRSP) and Compustat files. The patent database comes from the public information on Professor Leonid Kogan and Noah Stoffman’s webpage updated until the end of 2020.⁸ In addition, we collect the information of trade secret laws of the UTSA and IDD from Png (2017a, b), Glaeser (2018), Klasa et al. (2018), and Li et al. (2018). The UTSA and IDD information is at state-level and the year each state enacted and adopted the UTSA and IDD is shown in Table 1. Because state information in Compustat database is updated to the most recent location of the firm’s headquarters, actual historical headquarters information is not available in the database. We thus use the Augmented 10-X Header Data for historical location data, which is available since 1994 because the SEC EDGAR began requiring firms to upload their 10-K filings to EDGAR that year.⁹ Therefore, we restrict our IPO sample period from 1994 to 2020. Our final sample consists of 3,888 IPOs. Table 8 provides the detailed illustrations of the sources and construction of our data.

3.2 Variables and methodologies

To investigate our hypothesis that the adoption of trade secret law increases information asymmetry of IPO, we first follow Liu and Ritter (2011) and Loughran and Ritter (2004) and use the first-day returns of IPOs as the proxy for IPO underpricing. The first-day return of IPO (First-day return) is the percentage change between the initial public offering price and the first-day closing price of IPO.¹⁰ To examine the influence of trade secret laws, we adopt the dummy variable of trade secret laws, which is equal to one for firms located in states that either adopt the UTSA or IDD, and zero otherwise. In addition, we consider that the heterogeneity of trade secret protection laws from UTSA, IDD, and DTSA may have different effects on IPO underpricing. Thus, we also use separate dummy variables including UTSA and IDD, Only UTSA, Only IDD, and DTSA, to capture the different scenarios in which firms are located in states that adopt both UTSA and IDD, in states that only adopt UTSA, in states that only adopt IDD, and in states that adopt DTSA.¹¹

In the regression analyses, we follow Chambers and Dimson (2009) and Liu and Ritter (2011) and use firm size (Log(size)), IPO proceeds (Log(IPO proceeds)), book-to-market ratio (B/M), R&D intensity (R&D intensity) and venture capital backed dummy (VC) as control variables to explain the first-day returns of IPO. In addition, we consider the influence of industry and incorporate industry fixed effect into the regression. The implements of IDD and UTSA vary with states. To prevent the effects of IDD and UTSA on IPO underpricing from being influenced by state characteristics rather than legal protections, we also consider state fixed effects in the regression. Detailed definitions for all variables are given in Table 9.

3.3 Summary statistics

Figure 1 displays the annual average first-day return at the IPO day from 1994 to 2020. Over the past 27 years, we observe a positive first-day return in all years. The mean of the averaged first-day return of 27 calendar years is 23.7% (with t-statistic 6.49). The underpricing is more pronounced for the years in late 1990s, consistent with the finding in Ritter and Welch (2002) that IPO underpricing is more severe for the internet bubble period. We also find the evidence that IPO underpricing is increasing in the most recent 5 years. Additionally, we observe a positive correlation between IPO volume and underpricing throughout our study period.

Fig. 1

Average first-day returns at the IPO day and IPO volumes. The graph presents a line graph of averages of first-day returns on the IPO day and a bar chart of the IPO trading volumes from 1994 to 2020

Table 2 presents descriptive statistics of variables whereas Panel A displays summary statistics and Panel B displays the correlation matrix. In Panel A, the average (median) of first-day return of IPOs is 24.2% (11.3%). We examine the average first-day return of IPOs each year from the Ritter website.¹² By calculating the annual data of his reports during the period of 1994–2020 (that is, our sample period), we obtain the average first-day return of IPOs at 24.24%, which is almost the same as our results. In addition, studies of U.S. IPOs such as Liu and Ritter (2011) and Chen et al. (2022b) respectively show average IPO first-day return at 24% and 32.33% for the periods 1993–2008 and 1990–2016.¹³ Thus, the average first-day return results of IPOs in our paper are quite similar to those of previous research. Moreover, the mean of the legal dummy variable helps us to understand the proportion of IPOs under trade secret laws. We find that about 92% of IPO firms are located in states that either adopt the UTSA or IDD, and roughly 19% of IPO firms are located in states that both adopt the UTSA and IDD. We also find that a higher percentage of IPO firms are located in UTSA-only states than in IDD-only states. This finding as a proportion of IPOs is likely because more states have implemented UTSA than IDD (Table 2).

Table 2 Summary statistics and correlation coefficient matrix

The correlation matrix in Panel B displays the Pearson correlation coefficients between variables. We find a slight positive relationship between the first-day return and the trade secret laws, with a correlation coefficient of 0.0486. The first-day return is also positive correlated with other dummy variables of trade secret laws, except the dummy of UTSA and IDD. These findings appear to indicate a positive association between IPO underpricing and trade secret laws. There is a positive relationship between the first-day return and the log(size), implying that larger IPO firms may exhibit higher first-day returns. The correlation coefficient between log(size) and log(IPO proceeds) is about 0.81, indicating that larger IPO firms typically have greater IPO proceeds, consistent with previous research. In addition, there is a positive correlation between the VC dummy variable and IPO proceeds, indicating that venture capital backed IPO firms tend to have greater IPO offerings. Further, the first-day return demonstrates a positive correlation with the VC dummy variable, suggesting that firms supported by venture capital firms tend to experience higher first-day returns, as demonstrated by Liu and Ritter (2011).

4 Empirical results

4.1 IPO first-day returns under different trade secret laws

In order to preliminarily investigate the impact of trade secret protection laws on IPO underpricing, we compare the IPO first-day returns under different trade secret laws. Results are shown in Table 3. Panel A provides summary statistics of first-day return of IPOs categorized by trade secret laws. Specifically, we divide IPO firms into four groups: firms without UTSA and IDD protection, firms with either UTSA or IDD protection, and firms with both UTSA and IDD protection. The results of Panel A show that IPO firms in states with UTSA protection but without IDD protection and firms in states with IDD protection but without UTSA protection respectively have first-day returns of 26.6% and 25.3%. IPO firms with neither UTSA nor IDD protection have the lowest average first-day return, 14.3%. These results imply that the first-day returns of IPOs under the trade secret laws tend to be higher than those without trade secret laws.

Table 3 IPO first-day returns under different trade secret laws

Panel B of Table 3 presents the ANOVA and contrast tests to compare the significant differences across groups. The ANOVA test indicates a statistically significant difference in first-day returns among the four subsamples. The findings show that the three subsamples with trade secrets protections under UTSA and/or IDD have significantly higher average first-day returns than firms without trade secret laws. In addition, the contrast tests show that firms with both UTSA and IDD protection have significantly higher IPO underpricing than firms without any protection of trade secret laws. Further, firms protected only UTSA have first-day returns similar to those of firms protected only by IDD, suggesting that there is no significant difference in effectiveness of protection between these two different trade secret laws.

4.2 Regressions of IPO first-day returns

Table 4 presents the regression results of IPO first-day returns on trade secret laws. There are four models in Table 4. Model 1 is our baseline model and only examines the impact of trade secret laws on first-day returns of IPOs. Model 2 incorporates the industry and state fixed effects into baseline model.¹⁴ Model 3 additionally considers the effect of DTSA and the influence of firm characteristics, including Log(size), Log(IPO proceeds), B/M, R&D intensity and VC.¹⁵ Further, Model 4 decomposes the effectiveness of the trade secret protections into different scenarios of protection laws.¹⁶ In the models (3) and (4), we use the DTSA dummy variable as a supplement to state-level trade secret laws such as IDD and/or UTSA in protecting corporate innovation rather than replacing them because DTSA was federal law and was passed after most states had passed state-level protection laws. The related regression equations are:

$${\text{Model}}\,{1}:First\!-\!day \, return = \alpha + \beta_{1} Trade \, secret \, laws + \varepsilon ,$$

(1)

$$\begin{aligned} {\text{Model}}\,2:First\!-\!day \, return & = \alpha + \beta_{1} Trade \, secret \, laws + \delta_{1} Industry \, FE \\ & + \delta_{2} State \, FE + \varepsilon , \\ \end{aligned}$$

(2)

$$\begin{aligned} {\text{Model}}\,3:First\!-\!day \, return & = \alpha + \beta_{1} Trade \, secret \, laws + \beta_{2} DTSA \\ & + \gamma Control \, variables + \delta_{1} Industry \, FE + \delta_{2} State \, FE + \varepsilon , \\ \end{aligned}$$

(3)

$$\begin{aligned} {\text{Model}}\,4:First\!-\!day \, return & = \alpha + \beta_{1} UTSA \, and \, IDD + \beta_{2} Only \, UTSA \\ & + \beta_{3} Only \, IDD + \beta_{4} DTSA + \gamma Control \, variables \\ & + \delta_{1} Industry \, FE + \delta_{2} State \, FE + \varepsilon . \\ \end{aligned}$$

(4)

Table 4 Regression analysis of IPO first-day returns on trade secret laws

First, we consider the impact of trade secret protection laws. In Models 1 to 3 of Table 4, the coefficients of trade secret laws are significantly positive, showing that trade secret protection resulted from UTSA and/or IDD implements increases the first-day return of IPOs.¹⁷ In regression results, the effect of DTSA is not significant, possibly because all states had adopted IDD and/or UTSA before DTSA was passed in 2016. The significant protective effect of UTSA and/or IDD may reduce the effect of later implementation of DTSA. In addition, the impact of DTSA is less significant because it was adopted in the United States in 2016 and affects only a few years in our sample, 1994 to 2020.

Model 4 compares the effectiveness of protection from different trade secret laws of state level. The results indicate significantly positive coefficients of UTSA and IDD, Only UTSA, and Only IDD. These findings show that whether the trade secret law is based on UTSA or IDD, the positive impact of trade secret laws on IPO underpricing is significant.¹⁸ We show that the adoption of trade secret laws leads to information asymmetry and IPO underpricing, which indicates that trade secret law is beneficial for firms in protecting their know-how in the form of trade secrets, reducing the disclosure of information. Overall, these findings are consistent with our hypothesis that trade secret laws encourage firms to invest in innovations with trade secrets and increase information asymmetry, leading to increased underpricing of IPOs.

The influences of control variables from firm’s characteristics on the first-day returns of IPOs are shown as follows. First, the coefficient of VC are significantly positive, showing that VC backed IPOs have the significantly positive first-day returns. This result is consistent with previous studies such as Liu and Ritter (2011), Peng et al. (2021), and Yang and Yuan (2022).¹⁹ Second, R&D intensity has a significantly negative impact on the first-day returns of IPOs. This finding is inconsistent with previous research such as Guo et al. (2006) suggesting that firms with high R&D intensity tend to have higher uncertainty and information asymmetry. However, we still find a positive effect of R&D intensity on first-day returns when we begin the sample from 1970 and disregard firm headquarters information used to test the effect of trade secret laws. Thus, our negative result for R&D intensity is likely caused by the different time period of the sample. In addition, since firms with higher R&D intensity tend to have more trademarks, our negative R&D effect may be explained by Yang and Yuan (2022), who find that IPO firms with more trademarks tend to have less IPO underpricing.²⁰ Third, the firm size and IPO proceeds have inverse impact on the first-day return of IPO, a result similar to that of Yang and Yuan (2022). This finding may result from the strong relationship between firm size and IPO proceeds.

4.3 Robustness checks

In this subsection we perform robustness checks to prevent our results from being dominated by the influences of specific firms. Table 5 shows the robustness checks for the regression analysis of IPO returns on trade secret laws. First, Chen et al. (2020b) find that about 28% of R&D firms are located in California. In addition, 25.49% of our IPO sample is located in California. Because more IPOs issued in California tend to be technology firms, their high IPO first-day returns may be the result of technology uncertainty. To prevent our results from being driven by the subsample of IPOs in California, we follow the approach of Chen et al. (2020b) and remove IPO firms located in California. Results are shown in Panel A of Table 5.

Table 5 Robustness checks for the regression analysis

Second, we consider the literature on IPOs such as Ritter and Welch (2002) and Kim and Ritter (1999), which shows that penny stock and small firms tend to have high information asymmetry and are more likely to have significantly positive IPO first-day returns. To avoid the potential for such firms to skew our results, we exclude IPOs with stock prices of < 3 and small firms with total assets of < 10 million in Panels B and C of Table 5, respectively. Third, we also consider the endogeneity problem of location selection for IPO firms: the implementation of the trade secret law may affect the location selection of corporate headquarters. We thus follow Chen et al. (2020b) and perform a Heckman two-stage regression to control for the headquarters location choice. We show only the second stage results in Panel D.²¹

Overall, the results of the robustness checks shown in Table 5 show the significantly positive impact of trade secret laws on IPO first-day returns. These results are consistent with our main hypothesis and enhance the reliability and validity of our main findings, suggesting that the protection effect of trade secret laws increase the information asymmetry, leading to IPO underpricing even when controlling for various factors and subsamples.

4.4 Regressions considering the effects of private disclosure

Our empirical findings demonstrating the effect of trade secret laws on IPO underpricing shows that the legal protection of trade secrets enables firms to appropriate their innovations, leading to information asymmetry. Intuitively, trade secret laws are effective in providing protection only when a firm’s know-how is held in the form of trade secrets. To directly capture the trade secret effect of firms, we manually collect disclosures of whether firms possess trade secrets from 10-K reports. Ettredge et al. (2018) find that about 30% of firms disclose that they have trade secrets in the 10-K reports (without disclosing the technology details).²² Thus, we perform a further analysis using trade secret disclosures in 10-K filings (Glaeser 2018).²³

Table 6 presents the regressions considering the effects of private disclosure from 10-K filings. Model 1 is the baseline model and only examines the impact of firm-level trade secrets disclosed in 10-K on the first-day returns of IPOs. Model 2 incorporates the industry and state fixed effects into the baseline model. Model 3 additionally considers the effect of DTSA and the influence of firm characteristics, including Log(size), Log(IPO proceeds), B/M, R&D intensity and VC. Further, Model 4 decomposes the effect of firm-level trade secrets (i.e. 10-K trade secret) into two parts: protected by trade secret law and not protected by trade secret law. Specifically, in Model 4, the interaction term (A) is designed to capture the effect of firm-level trade secrets under the protection of trade secret laws, whereas term (B) represents the impacts in the absence of such legal protection.

Table 6 Regression analysis of IPO first-day returns on 10-K trade secret disclosures

In Models 1–3 of Table 6, the coefficients of 10-K trade secret disclosure are significantly positive, showing that a firm’s disclosure of trade secrets in its 10-K report increases its IPO first-day returns. These results suggest that preserving know-how in the form of trade secrets makes investors less able to accurately assess a firm’s intrinsic value, leading to significant IPO underpricing. In addition, Model 4 examines and compares the effect of 10-K trade secrets on IPO first-day returns with and without trade secret protection. The results indicate significantly positive coefficients of both interaction terms (A) and (B), showing that whether or not a trade secret law exists, the positive impact of a firm’s trade secrets on IPO underpricing is significant. Specifically, even without the protection of trade secret laws, preserving proprietary technology in form of trade secrets within the firm still causes IPO underpricing and thus have a positive impact on IPO first-day returns. A test of the coefficient difference between (A) and (B) shows that the first-day IPO returns of firms that disclose trade secrets and are protected by trade secrets are significantly higher than those of firms that disclose trade secrets but are not protected by trade secrets. This finding shows that the existence of the trade secret law increases the protective effect of firms keeping their know-how in the form of trade secrets, thereby strengthening the information asymmetry between investors and firms, leading to more IPO underpricing.

4.5 Moderating effects of trade secret laws on IPO underpricing

Next, we study moderating effects of trade secret laws on IPO underpricing, including roles of covenant not to compete (CNC) agreements, complex industry, patent systems, R&D intensity, and firm size. For the moderating effects, we adopt a similar approach, first dividing the sample into two groups by using the interaction terms in the regression. Second, we use the test of coefficients difference between the interaction terms (A) and (B) to identify key factors contributing to the moderating effects. Table 7 presents the empirical results of the analysis of moderating effects, where the dependent variable is the first-day return of the IPO firm.

Table 7 Moderating effects of trade secret laws

Model 1 of Table 7 examines the relationship between trade secret laws and IPO underpricing for IPO firms located in states strong and weak CNC agreement. A CNC is a legal agreement that restricts former employees from working for a competitor within a specific time or distance, helping firms indirectly protect their confidential information (Gilson 1999). Trade secret laws and CNC agreements may be alternative tools for employers to choose to sue for violations of use of confidential information. In this study, we define strong CNC as a binary variable with a value of one if a firm is located in a state with a strong CNC index, meaning that its Garmaise’s (2011) CNC index is higher than the median of the CNC index, and zero otherwise. Weak CNC is defined as one minus strong CNC. Our results show that the underpricing of IPOs is significant for firms in states with trade secrets but with weak CNC, while it is not significant in states with both trade secret laws and strong CNC. Our empirical results show that CNC and trade secret laws may to some extent be substitutable in protecting intellectual property rights.

Model 2 investigates the relationship between trade secret laws and IPO underpricing, taking into account industry complexity. Cohen et al. (2000) propose that firms in complex product industries use patents to force competitors to negotiate, while firms in discrete product industries use patents to prevent competitors from developing substitutes. Following Cohen et al. (2000), we define a dummy variable, more complex industry, with a value of one if a firm operates in complex industries such as telecommunications equipment or semiconductors, and zero otherwise. Less complex industry is defined as one minus more complex industry. This allows us to investigate the potential influence of industry complexity on the correlation between IPO underpricing and trade secret protection. We find that trade secret law has a significant impact on IPO underpricing in both more and less complex industries. Additionally, the Chi-square statistic indicates that IPO firms in more complex industries experience greater underpricing than IPO firms in less complex industries. These findings suggest that the effectiveness of trade secret protection varies with industry complexity, and firms operating in complex industries with trade secret protections may face more significant information asymmetry during the IPO process.

Model 3 examines whether patent systems may affect the relationship between IPO underpricing and trade secret laws. We measure a dummy variable, more patents, with value of one if a firm is located in a state with an average number of patents greater than the median, and zero otherwise. Fewer patents are defined as one minus more patents. The dummy captures to what extent a state relies on a patent system to protect its intellectual property rights. The results show that both coefficients of (A) and (B) in Model 3 are similar and significantly positive, demonstrating that IPO firms with trade secret laws have similar IPO underpricing whether the state has more or fewer patents. The Chi-square value also confirms that they are not statistically different.

Model 4 studies the impact of whether firms have R&D spending on the relationship between IPO underpricing and trade secret law. We construct a dummy variable, with R&D expenses, which is with value of one if a firm incurs R&D expenses, and zero otherwise. Without R&D expenses is defined as one minus with R&D expenses. We find that for firms headquartered in states with trade secret laws (whether with or without R&D expenses), there is significant underpricing of IPOs. In addition, the Chi-square value reveals that firms with trade secret protection and R&D expenditures appear to experience stronger IPO underpricing, indicating that trade secret legal protection may have a more significant effect on IPO underpricing for firms with R&D activities.

Model 5 evaluates the impact of firm size on the relation between IPO underpricing and trade secret legal protection. We examine the effect of firm size, because firm size affects the degree of information asymmetry of IPO firms to a certain extent. Small firms, which have greater information asymmetry, may experience severe IPO underpricing, making trade secret laws less effective. In Model 5, we find that both large and small IPO firms located in a state with trade secret laws are prone to significant IPO underpricing. The Chi-square value suggests that the coefficient difference of trade secret laws between large and small IPO firms is significant, showing that the effect of trade secret laws is more important for large IPO firms.

These results have the following implications. First, the innovative technologies of firms in complex industries such as telecommunications equipment and semiconductors are more difficult to understand, thereby increasing the protective effectiveness of trade secret laws. In addition, considering that R&D is frequently regarded as an intangible element contributing to a firm’s valuation, trade secrets owned by firms actively engaged in R&D activities are likely to have greater value. Therefore, R&D investments strengthen the impact of trade secret laws on IPO first-day return. Third, the additional information asymmetry problem created by trade secret laws prohibiting the disclosure of confidential information is more severe and sensitive to large firms because these firms have greater transparency than smaller firms. Thus, there is an enhanced incremental effect of trade secret laws on large IPOs. Our findings suggest that trade secret laws increase the difficulty investors face in assessing the intrinsic value of IPO firms operating in complex industries, IPO firms with R&D investments, and large IPO firms.

5 Conclusions

This study explores the impact of trade secret laws on IPO underpricing in the U.S. We hypothesize that the unique nature of trade secrets, which entail limited transparency, creates a higher degree of information asymmetry that complicates the valuation of the IPO firm, leading to greater IPO underpricing. We examine the influences of two trade secret laws: the UTSA and IDD. Our empirical results show that IPO underpricing is more pronounced in states with trade secret laws than in states without trade secret laws. Robustness checks based on excluding California-based firms, penny stocks, small firms, and controlling for location choice issue still obtain consistent results. In addition, using disclosure information of whether the firm has trade secrets in the 10-K report for further analysis also confirms this finding. Finally, we study the moderating effects of trade secret laws on IPO underpricing by examining covenant not to compete (CNC) agreements, complex industry, patent systems, R&D intensity, and firm size. We show that the effects of trade secret laws are more pronounced for firms operating in a more complex industry, firms with R&D expenses, and larger IPO firms.

The findings of this study have important implications for both entrepreneurs and investors. When a firm goes to the IPO process, entrepreneurs must weigh the benefits from protection of trade secret laws and the potential costs of IPO underpricing. In addition, entrepreneurs working in companies located in states with trade secret laws, particularly in complex industries, companies with R&D expenses, and larger companies, should pay more attention to the potential occurrence of IPO underpricing when the company is preparing to go public. Entrepreneurs may be attentive to the potential influence of trade secret legal protection on IPO underpricing and take proactive steps to minimize the risks associated with information asymmetry. For example, entrepreneurs can protect some of their intellectual property rights using the patent system instead of trade secrets. Moreover, investors should recognize and consider the potential effects of trade secret protection on IPO underpricing when valuing and assessing investment opportunities. This may involve incorporating additional risk factors and adjusting expectations regarding the initial pricing of IPOs from firms located in states with robust trade secret laws.

Appendix

See Tables 

Table 8 Data source and collection procedure

8 and

Table 9 variable definitions

9.