Table of contents
Data and Method
Silicon Valley is home to 6 out of the 10 most valuable companies in the world and is considered to be the worldwide leader in venture capital. However, at the same time, the area faces the highest living costs and business expenses in the nation. Reports suggest that people and businesses have been leaving the area due to the COVID-19 pandemic. This study takes a time series analysis approach in order to determine whether there has been a structural change in Silicon Valley's startup funding activity during the COVID-19 pandemic. A dataset of 13,350 funding rounds in Silicon Valley from 2016 to 2020 has been collected from Crunchbase. The time series analysis is conducted in R using the Strucchange package and Chow tests. Results suggest that no structural change has occurred in the total startup funding activity or in the proportion of early-stage investments during the COVID-19 pandemic.
Prominent startup ecosystems are the birthplace ofnew unicorn startups every month, with Silicon Valley creating as many as 28 new unicorns just over October and November in 2020 (Costa, 2021). A startup reaching unicorn status means that the private company has achieved a valuation of at least $1 billion, which has become the standard objective and status symbol for high-growth startups (Jinzhi & Carrick, 2019). The start of 2021 saw an all-time high of 701 worldwide unicorns, 182 of which were new that year (Savitz, 2021). These unicorn companies can be found in entrepreneurial ecosystems spread over 35 countries, 52% of these unicorns are in the US and 21% in China, with examples including SpaceX, Stripe, and ByteDance (Savitz, 2021).
Entrepreneurial or startup ecosystems are defined by existing literature as “a collection of entrepreneurs creating companies and innovative products and services for many users and agents in the global economy” (Sussan & Acs, 2017). The main characteristics of these entrepreneurial ecosystems are the concentration of talent, knowledge, and access to capital (Engel, 2015). These ecosystems take the form of areas that house a significant number of startups, supporting businesses, and mature companies, with the most famous example being Silicon Valley (Engel, 2015).
Silicon Valley became the center of semiconductor production in the 20th century, being home to 26 out of all 31 US semiconductor manufacturers in 1960 (Dennis, 2019), which attracted tremendous capital and talent to the region. Silicon Valley continuously adapted when personal computers came along in the ‘80s and ‘90s, and with computer software and internet-based companies after that (Dennis, 2019). Today, Silicon Valley is the worldwide leader in terms of venture capital and successful startups, and is home to 6 out of the top 10 most valuable companies in the world (Haqqi, 2021). The area is known as a ‘startup heaven' as a result of the many opportunities, community, and mentality that can be found there (Shobith, 2021). Access to venture capital is a significant part of the reason that startups and entrepreneurs from around the world flock to Silicon Valley, as research suggests that venture capital has a direct positive influence on a startup's ability to succeed (Chang, 2004; Jeong et al., 2020).
However, at the same time, Silicon Valley faces the highest living costs, state taxes, and business expenses in the nation (Brinklow, 2019). This raises the question of whether it remains worth it for startups to remain in, or move to Silicon Valley. Speculators and initial reports suggested that as a result of the COVID-19 pandemic, tech talent and venture capitalists were leaving Silicon Valley in large numbers (Ioannou, 2021; Lindzon, 2020). Given the importance of venture capital for the ecosystem, it is imperative to have a clear overview of the changes to venture capital activity and the impact that this has on startups located or interested in the region.
This research takes a time series analysis approach to examine what has happened to the venture capital activity in Silicon Valley over the past years. Additionally, the research attempts to examine the impact of COVID-19 on startup funding activity. Given that the COVID-19 pandemic is a recent phenomenon, there are only a limited number of articles, research papers, and databases available to use for this research. This study considers all startup funding rounds that took place in Silicon Valley during a 5-year period between 2016 and 2020. Exclusive academic access for Crunchbase was obtained for this study, which is a database containing information about startups around the world and is considered to be “the largest public database with profiles about companies” (Liang & Yuan, 2016). This study is centered around the following research questions: “What influence has the COVID-19 pandemic had on Silicon Valley's startup funding activity?”.
The next section of this paper contains the literature review, where the existing literature is evaluated and the hypotheses are introduced. The section after that describes the data, methods, and analyses that are used in the research. The results of the analyses, and an initial interpretation, are shown in the next section. Following the results section, an extensive discussion will relate this study's results to the existing literature, outline the limitations, and suggest points for future research. Finally, the paper finishes with a conclusion summarizing the research and its contribution.
As was briefly introduced in the previous section, entrepreneurial ecosystems are defined as a cluster of entrepreneurs that create innovative services and products (Sussan & Acs, 2017). The ecosystems are characterized by a concentration of talent, knowledge, and access to capital (Engel, 2015). Cukier et al. (2016) take a more geographical approach, as they define an entrepreneurial ecosystem as a region within one hour of travel range (or 30 miles), consisting of startups and a variety of supporting organizations, that work together in order to create new startups and further develop existing startups.
Entrepreneurs are faced with the decision of where to locate their business, which presents itself to be complicated due to the many factors and trade-offs involved. One of those trade-offs is between whether entrepreneurs are better off remaining in their hometown, where they are socially embedded, or moving to an entrepreneurial ecosystem where more resources are available (Guzman, 209). Research considering startups migrating to Silicon Valley suggests that the access to plentiful resources in a large ecosystem is more significant for a startup's performance than its social embeddedness (Guzman, 2019). Guzman's (2019) research indicates that migration has an impact on startup performance. A study conducted by Chung et al. (2020) agrees with Guzman on the importance of a startup's location on their success, however, takes it a step further by differentiating between the different kinds of ecosystems. The article describes how the government, private companies, and universities all contribute to the development and performance of ecosystems. The Triple Helix model, on which the article is based, describes that a balance of the three influences is needed for an ecosystem to become successful (Etzkowitz, 2003). Chung et al. (2020) suggest that ecosystems that have a stronger government influence yield the highest performance, which is due to government policies that support the development of new technology. However, this study was conducted in South Korea, using data from local entrepreneurial ecosystems. Since there are no South Korean areas in the top 20 worldwide ecosystems (ranked by total venture capital investments), and with cultural and business-climate differences, it is possible that the conclusion from this study is not representative of other areas like Silicon Valley (Florida & Hathaway, 2018).
In a 2018 research article, Pique et al. examined Silicon Valley using the Triple Helix model and concluded that the influence of the triple helix model on startups in Silicon Valley depends on the phase that startups are in. As shown in image 1 below, the article concludes that some components of the model have a more positive or negative impact depending on whether the startup is in the inception, launch, growth, or maturity phase. This contradicts Chung et al. (2020)'s findings, by suggesting that there is not one single component of the Triple Helix model that is most important.
Importance of Triple Helix model components to startups in Silicon Valley (Pique et al., 2018).
Abbildung in dieser Leseprobe nicht enthalten
The conclusions of the aforementioned research articles demonstrate that the location of a startup is in fact important and should be carefully considered. The articles show that the characteristics of a startup ecosystem have an impact on the success of startups. Depending on which stage the startup is in, it might for example benefit more from university resources, such as readily available talent, rather than government resources, such as grants or subsidies. An ecosystem like Silicon Valley provides access to a wide range of resources, such as talent from world-class universities like Stanford, which other or smaller ecosystems may not be able to provide. Additionally, Silicon Valley also has a long history of government support in the form of venture capital incentives (Golomb, 2014), is home to 13.5% of the world's startup deals (Dautovic, 2020), and is home to 6 out of the top 10 most valuable companies in the world (Haqqi, 2021), indicating that it scores well across the board with regards to the Triple Helix model.
Another study, conducted by Engel (2015), examined what helped transform Silicon Valley from a small agricultural area to the global tech powerhouse it is today. Engel (2015) found that three key components contributed to this transformation, which are universities, government, and entrepreneurs. The universities in the area, such as Stanford and the University of California Berkeley, initially started with a focus on agricultural education. However, in the 1900s, the universities took notice of the innovation that was happening in the telecom, computer technologies, and electronics industry, and shifted their focus to technology and business education. Since then, Stanford has been at the forefront of advanced research and is ranked as the 3rd best university worldwide (Nietzel, 2021).
As was previously mentioned, the government also had a large influence through its grants and subsidies. However, according to Engel (2015), the government also funded military research, such as physics at Berkeley and electronics at Stanford, and funded research laboratories. Military contracts also helped build some of the pillars of Silicon Valley, such as Hewlett Packard and Lockheed. Engel (2015) argues that government spending was a crucial driver for the subsequent emergence of Silicon Valley as a technological innovation ecosystem.
The third and final component of Engel's (2015) study is the entrepreneurs in Silicon Valley. Engel (2015) describes how there is an entrepreneurial spirit in Silicon Valley that evolved from the California Gold Rush days, where individuals would venture out and take big risks in hope of receiving big rewards. This spirit continues today, in the form of entrepreneurs willing to take a significant risk by taking on venture capital in order to pursue potential significant upsides. Engel (2015) also mentions how this entrepreneurial spirit leads successful entrepreneurs to stay engaged within the ecosystem, for example, after founders reach an exit they will often stay involved by investing in, or giving advice to, other startups.
These components are similar to the ones considered in the Triple Helix model, however, Engel (2015) places more emphasis on the importance of an entrepreneurial community, since entrepreneurs not only start the companies, they often also recycle themselves after exits, and the entrepreneurs further contribute to the culture and spirit of Silicon Valley.
The year 2020 has been anything but normal due tothe COVID-19 pandemic, and it is thus imperative to examine whether the pandemic has had an effect on the importance of location for startups. While the literature agrees on the advantages of entrepreneurial ecosystems (e.g. Tripathi et al., 2019), in 2020 the world has seen major changes as a result of the COVID-19 pandemic. The pandemic has changed the way we work, as people around the world have been working remotely and from home. This shift to working from home could potentially also have long-term effects, as companies and employees are voicing their dissatisfaction with the daily 9-to-5 office culture. A global survey conducted by Slack reflects this dissatisfaction, showing that only 12%of employees want to return to full-time office work and 72% of respondents suggest moving forward with a hybrid model, being a combination of working from home and some time inthe office (Slack, 2020). A range of major tech companies, including Facebook and Google, have already stated their support for another year or two working from home, with companies like Twitter and Square taking it a step further by stating that their employees could work from home indefinitely (Liu, 2020).
This shift from working in the office to working from home has led to companies downsizing or even completely disbanding their offices, with Silicon Valley seeing its office vacancy rates rise to 16.7 percent, the highest ina decade (Bowles, 2021). At the same time, Silicon Valley's residential rent has decreased 27% since last year and the number of homes for sale has increased (Bowles, 2021), indicating that people are moving out of the area. As is similar within the other major startup ecosystems, the cost of living in Silicon Valley is significantly higher than the US national average, in fact, Silicon Valley has the highest national living cost (Brinklow, 2019). With employees working remotely, it bears the question of whether people should continue to live in an area where their living costs are exorbitantly high. The same goes for companies, whether a startup or mature, which are paying a premium in terms of rent and wages to be located within a startup ecosystem like Silicon Valley.
When considering the aforementioned, it is apparent that there has been an outflow of people and companies from Silicon Valley, and it is thus imperative to analyze the impact that this has had on the ecosystem. While it is not possible to forecast whether people and companies will return to (or remain in) Silicon Valley once the COVID-19 pandemic is fully over, it is possible to examine the impact that the pandemic has had so far by analyzing the venture capital data of the area. Recent reports give a preliminary indication of venture capital activity, as statistics show that Q3 2020 venture capital investments reached the second-highest levels ever in the US (Costa, 2021). The unprecedented times as a result of the pandemic and the significance of such reports indicate that there could be a structural change in the levels of startup funding. This would mean that there is an abrupt change in the time series, such as a sudden and significant change in the startup funding activity. In order to test whether the startup funding activity in Silicon Valley has structurally changed during the COVID-19 pandemic, this paper proposes the following hypothesis:
H1: Silicon Valley's startup funding has structurally changed during the COVID-19 pandemic in 2020.
Venture capital is a method of financing that takes the form of a company exchanging equity or debt in return for capital. In its simplest form, the venture capitalist invests in a startup, supports it for a period of time, and eventually exits during a liquidation event in order to receive a return on their investment (Zider, 1998). Venture capital has become an increasingly more popular choice for startup funding over the past 30 years, with the number of active US venture capital investment firms growing from 408 in 1991 to 1,639 in 2015 (Hong et al, 2020). During that same period, the number of US companies that received venture capital investments increased from 970 in 1991 to 3,743 in 2015 (Hong et al, 2020).
Venture capital firms are usually structured as limited partnerships, with the general partners being the individuals that manage the investments, and the limited partners being the individuals or funds supplying the capital (Sarkissian, 2017; Zider, 1998). Limited partners typically include pension funds, foundations, and insurance companies, and will usually pay an annual management fee between 1 and 3 percent of the total committed capital (Sarkissian, 2017). Additionally, the general partners typically retain 20% of the fund's net gains which acts as a reward, but also as an incentive to keep the general partners' interests aligned with that of the limited partners (Sarkissian, 2017). The average lifespan for a venture capital fund is between 8 to 12 years, during which the firm will enter into new deals and expect their existing deals to exit (Wagner, 2014). Examples of prominent venture capital firms include Kleiner Perkins, Accel, and Sequoia Capital (Jain, 2020).
Venture capital investments are separated into types of rounds, depending on the development phase or level of maturity that the company is in. The first investment type is the seed round, which is the most common investment type for early-stage startups and is primarily used to develop the product and business (Tripathi et al, 2019). In 2018, startups in Silicon Valley raised an average of $5.6 million during the seed round (Loizos, 2019). After successfully raising a seed round, startups will continue to develop their product, gain proof of concept, and gather metrics, which the startup can use to raise the next round of financing; series A (Cremades, 2018). In 2018, startups in Silicon Valley raised an average series A round of $15.7 million (Loizos, 2019). Once the startup has proven that its business model works and for example has the intention to scale through expansion or acquisition, the startup can consider a next investment round: the series B. In 2018, startups in Silicon Valley raised an average series B round of $30.7 million (Loizos, 2019). As the startup grows further and requires more venture capital, subsequent investment rounds can be organized until the company no longer requires external capital or reaches the point of an exit.
An exit for a company, also known as a liquidation event, can take the shape of either a merger and acquisition (M&A) or an initial public offering (IPO) (Berry, 2017). During an M&A, the company will either merge together with or be acquired by, another organization which involves the shareholders selling some or all of their shares. On the contrary, during an IPO the company will ‘go public', meaning that the shares of the company will be publicly tradable on an exchange. Once the shares are publicly tradable, shareholders are able to sell their shares and have an exit, or hold on to their shares and sell at a later moment. However, initial public offerings are significantly less common than mergers and acquisitions, due to the lengthy and expensive process of taking a company public. This becomes clear when examining that 97% of exits in 2016 took the form of mergers and acquisitions (Lunden, 2017).
Venture capital is the most popular source of funding for startups, as the high-risk nature of startups often makes them unsuitable for traditional business finance such as bank loans (Bettignies & Brander, 2006; Tripathi et al., 2019). The main difference between the two is that a bank loan allows the founders to retain ownership, whereas a venture capitalist will take equity in exchange for the capital (Bettignies & Brander, 2006). Overall, venture capital is considered to be a key influence in the growth of startups, value creation, and innovation (Bocken, 2015). A study considering US startups found that venture capital investments, especially in the early stage of a startup, lead to sustainable high growth (Jeong et al., 2020). Tripathi et al. (2019) agree on the importance of venture capital, as the study concludes that the creation of startups can be negatively influenced by a lack of access to funding. The study continues by stating that the number of actively invested venture capital funds is a predictor for the size of a startup ecosystem (Tripathi et al., 2019). However, the benefits of venture capital go beyond the monetary value. A study conducted by Davila et al. (2003) examined the relationship between venture capital and startups in Silicon Valley using signaling theory and found that venture capital funding events act as significant signals about the startup's quality. Additionally, the same study also found that venture capital investments make it easier for startups to hire employees and grow their team, not just because of the capital, but especially because of the credibility that the signal creates (Davila et al., 2003). Another non-monetary benefit, which sets venture capital apart from traditional banking finance, is the managerial contributions of the venture capitalists (Bettignies & Brander, 2006). For example, a managerial contribution could come in the form of expert advice or access to a large network of potential employees and customers.
While the aforementioned literature agrees on the importance and influence of venture capital, setting up and closing an investment round presents itself to be a difficult task for startup founders. Venture capitalists regularly receive large quantities of proposals, and thus the general partners in charge of the fund have to go through a selection process in order to find the right opportunities for them. This process typically starts with the founders requiring an introduction to the investor, followed by a pitch from the founders explaining why their startup would be a good investment (Dennis, 2019). This shows that personal contact is central to the way of doing business in Silicon Valley. Dennis (2019) elaborates further on the importance of personal contact, stating that the personality of the founder and their personal presentation is a key determinant of whether a startup receives funding.
Interestingly, a study conducted by Lutz et al. (2013) found that the spatial proximity between the investor and startup has a significant impact on the likelihood of receiving funding. The study concluded that the probability of a founder having a financing relationship with an investor decreases as the distance between them increases (Lutz et al., 2013). While this study was conducted in Germany and the conclusions could possibly not be representative of Silicon Valley due to cultural differences, it could be an explanation for why the majority of large venture capital funds are located together on Sand Hill Road in the center of Silicon Valley. Another study reinforces the results from Lutz et al. (2013), as they concluded that the distance between startup and investor not only influences the probability of the financing relationship existing, it actually also influences the terms of the financing deal (Bengtsson & Ravid, 2009). Bengtsson & Ravid (2009) suggest that as the distance between the investor and startup increases, the investor is not able to monitor the startup as closely and will attempt to mitigate this risk through more contractual terms or clauses that work in favor of the investor. The conclusions from the previously mentioned literature indicate that the location of a startup is important with regard to venture capital.
However, employees have been working from home asa result of the COVID-19 pandemic and reports indicate that this has led toa significant number of people and businesses moving out of Silicon Valley (Ioannou, 2021). The area has the nation's highest living costs and business expenses, which incentivized people to relocate to lower-cost cities as they did not have to be physically present in Silicon Valley (Brinklow, 2019; Lindzon 2020). The outflow of people and businesses from Silicon Valley has likely created distance between startups and investors. Considering the literature from Lutz et al. (2013) and Bengtsson & Ravid (2009), this would negatively influence the startups' likelihood of successfully acquiring venture capital. Additionally, the distance would also make personal contact between the founders and investors more difficult, and as a result of the pandemic, there were little to no networking events where founders and investors could have met each other. Reports suggest that the economic insecurity as a result of the COVID-19 pandemic has led venture capital funds to invest less capital into risky, early-stage startups and instead invest more into proven, later-stage startups (Jungreis, 2020). Additionally, later-stage startups already have a network of existing and potential investors in place, which is something that an early-stage startup does not have. Thus, in 2020 it appeared to be more difficult for founders to meet investors, less likely to successfully raise venture capital, and investors reportedly moved more capital towards later-stage startups. In order to test the aforementioned, this paper proposes the following second hypothesis:
H2: The proportion of venture capital raised in early-stage rounds has structurally changed during the COVID-19 pandemic in 2020.
Data and Method
This study considers a dataset of 13,350 funding rounds in Silicon Valley over a 5-year period from 2016 to 2020. The data has been collected from Crunchbase, which is a crowdsourced database containing information about startups around the world and is considered to be “the largest public database with profiles about companies” (Liang & Yuan, 2016). The database contains records of companies, individuals, investment rounds, and liquidation events. The contents of the database originate from two main sources, community contributors and a large network of investors (Dalle et al., 2017). Community contributors are able to update the profiles of companies and individuals, while a network of global investors regularly updates their investment portfolios on the platform. Crunchbase has developed advanced algorithms that not only check the validity of the data but also enrich the data by crawling the internet for related information such as news articles or company announcements. Access to Crunchbase's database is restricted to their enterprise customers and selected universities, and thus exclusive academic access had to be obtained for this study. Crunchbase's data can only be accessed through their API by using JSON calls. One of the JSON calls used to collect data from the API is shown in table 1 below, and a full list of the JSON calls used in this study is shown in Appendix A for 2020, Appendix B for 2019, Appendix C for 2018, Appendix D for 2017 and Appendix E for 2016.
The example below shows the retrieval of all funding rounds in January 2020 within the location code ‘eb879a83-c91a-121e-0bb8-829782dbcf04', which represents the state of California. Every record represents a funding round and contains all of the associated properties as shown under ‘field_ids', which include a unique identifier of the funding round, the date on which the funding round was announced, the name of the funded organization, the location of the funded organization, the amount of money raised, the number of investors involved in the deal, and the type of investment.
Example of a Crunchbase API call used in the data collection.
Abbildung in dieser Leseprobe nicht enthalten
The Crunchbase API returns the data in JSON format, which is then converted into CSV format so that Alteryx can process it. Alteryx is a software program that is popular in the data science and analytics field, due to its big data analytics features (Berka, 2018). The first step in Alteryx was to group together all of the data by the respective year, as each month of data was collected separately. The data from the API was already relatively clean, however, there were still some blank spaces and trailing/leading whitespaces. The data cleansing function in Alteryx removed trailing/leading whitespaces and replaced blank spaces with ‘empty' or 0, depending on if the field is a string or numerical. The collected data contains funding rounds from California as a whole, and thus the next step is to filter out the funding rounds that took place within Silicon Valley. The study considers 72 cities that are all located in the Santa Clara Valley, as large as San Francisco or as small as Sonoma. A full list of the cities considered to be part of Silicon Valley is shown in table 3, which ranges all the way from Monterey to Santa Rosa.
List of cities included in Silicon Valley.
Abbildung in dieser Leseprobe nicht enthalten
A significant number of funding rounds did not have a funding amount listed on Crunchbase. Startups may choose to not disclose the amount of funding for a variety of reasons, however, since these funding rounds do not add value to our analysis, 6,541 funding rounds have been excluded. In February 2012, Broadcom secured a record $100B in post-IPO debt for a hostile takeover of Qualcomm. This funding round has been removed from the dataset, as it is considered to be an extreme outlier. Additionally, Broadcom solely secured the funding for the takeover, which ended up being blocked by the White House and thus never actually took place (Aiello, 2018). An overview of the Alteryx workflow reflecting all of the aforementioned steps can be seen in Appendix F.
After preparing the data, the final list of measures is as follows: funding round identifier, funded company name, funding round date, funding value, number of investors, city, and investment type. The investment type column contains information about what kind of funding round it was, such as ‘seed' or ‘series_c'. This can be used to determine whether the funding round can be classified as early-stage or late-stage. The round is considered early-stage if it is of the investment type ‘pre_seed' or ‘seed', and all other investment types are considered to be late-stage. It is worth mentioning that the investment type column also includes other types such as ‘convertible_note' or ‘series_unknown'. However, this does not have an impact on the validity of the analysis or comparison, as the same approach has been taken for all years
This study takes a time series analysis approach in order to analyze whether there has been any structural change in the startup funding data. The analysis is performed in R Studio using the R programming language. In order to test whether the time series contains any structural change, the Strucchange package from Zeileis et al. (2001) has been used. This package allows for the discovery of structural changes using the ‘sctest()' and ‘strucchange::breakpoints()' function, and the subsequent testing of breakpoints via a Chow test using the ‘sctest(type = "Chow")' function. The R language and Strucchange package have been used by a variety of similar studies (e.g. Bernabucci et al., 2015; Requier et al., 2015).
A descriptive overview of the results is shown in table 4 below. As becomes immediately apparent, there has been an increase in the total funding amount from 2019 to 2020, from $70.1 billion to $91.6 billion (30.7% increase). As would be expected with a higher total funding amount, the average funding amount also increased from $25.6 million to $36.9 million (44.1% increase). However, it is mention-worthy that even though the total and average funding amount increased, the number of rounds actually decreased from 2019 to 2020 (-9.5% decrease). Additionally, it also appears that there was a decrease in the total funding from 2018 to 2019, together with a decrease in the average funding amount and a decrease in the total number of rounds. Table 4 also lists the proportion of early-stage versus late-stage rounds per year. The results show a minor increase from 2018 to 2019 and another minor increase from 2019 to 2020.
Funding overview per year.
Abbildung in dieser Leseprobe nicht enthalten
Once the data set was loaded into R, the first step was to transform the data into a time series object using the line of code shown in Table 5. The time series is shown in Plot 1, which is an overview of the total invested capital per month over the full 5 year period. The upward-sloping blue line represents the trend line and is an indication of a long-term positive trend.
Creating the time series object.
Abbildung in dieser Leseprobe nicht enthalten
Total startup funding over time.
Abbildung in dieser Leseprobe nicht enthalten
Plot 2 below shows the same time series as in Plot 1, but then decomposed into observed, estimated trend, estimated seasonal, and estimated random categories. From this plot it can again be observed that there appears to be a long-term trend of positive growth in startup funding, however, the trend line seems to have turned negative in 2019 and recovered towards 2020. The plot also shows that the time series deals with some degree of seasonality, as is indicated by the seasonality pattern
Total startup funding over time, decomposed.
Abbildung in dieser Leseprobe nicht enthalten
Using the Strucchange package from Zeileis et al. (2001), a supF test was performed to determine whether structural change exists within the time series. This test considers the following H0: there is no structural change in the series. The lines of code used to perform the supF test are shown in Table 6 below. The test yields the following results: F = 16.351, p-value = 0.001. The p-value of less than 0.01 allows the H0 to be rejected and indicates that structural change exists within the time series.