资源描述
BIS Working Papers No 901 Inside the regulatory sandbox: effects on fintech funding by Giulio Cornelli, Sebastian Doerr, Leonardo Gambacorta and Ouarda Merrouche Monetary and Economic Department November 2020 JEL classification: G32, G38, M13, O3. Keywords: fintech, regulatory sandbox, startups, venture capital.BIS Working Papers are written by members of the Monetary and Economic Department of the Bank for International Settlements, and from time to time by other economists, and are published by the Bank. The papers are on subjects of topical interest and are technical in character. The views expressed in them are those of their authors and not necessarily the views of the BIS. This publication is available on the BIS website (bis). Bank for International Settlements 2020. All rights reserved. Brief excerpts may be reproduced or translated provided the source is stated. ISSN 1020-0959 (print) ISSN 1682-7678 (online)Inside the Regulatory Sandbox: Eects on Fintech Funding G. Cornelli S. Doerr L. Gambacorta O. Merrouche November 3, 2020 Abstract Policymakers around the world are adopting regulatory sandboxes as a tool for spurring innovation in the nancial sector while keeping alert to emerging risks. Using unique data for the UK, this paper provides rst evidence on the eectiveness of the worlds rst sandbox in improving ntechs access to nance. Firms entering the sandbox see a signicant increase of 15% in capital raised post-entry, relative to rms that did not enter; and their probability of raising capital increases by 50%. Our results suggest that the sandbox facilitates access to capital through two channels: reduced asymmetric information and reduced regulatory costs or uncertainty. Our results are similar when we exploit the staggered introduction of the sandbox and compare rms in earlier to those in later sandbox cohorts, and when we compare participating rms to a matched set of comparable rms that never enters the sandbox. JEL Codes: G32, G38, M13, O3. Keywords: ntech, regulatory sandbox, start-ups, venture capital. Giulio Cornelli (giulio.cornellibis), Sebastian Doerr (sebastian.doerrbis) and Leonardo Gambacorta (leonardo.gambacortabis) are at the Bank for International Settlements, Monetary and Economic Department. Ouarda Merrouche is at Universit e Paris Nanterre and EconomiX CNRS UMR 7235, 200 avenue de la R epublique 92001 Nanterre cedex. (ouarda.merroucheeui.eu). We would like to thank Matteo Aquilina, Raphael Auer, Marcel Bluhm, Nick Clark, Jon Frost, Luigi Guiso, Ivo Jenik, Daphn ee Papiasse and Manos Schizas, as well as participants at the BIS internal seminar and the BIS Innovation Hub. The views expressed here are those of the authors only, and not necessarily those of the Bank for International Settlements. 11 Introduction The rapid growth of innovative companies that use new technology (so-called ntechs) has the potential to transform the nancial sector fundamentally. 1 Fintechs hold the promise of spurring competition, leading to sizeable eciency gains, more choice for consumers and enhanced nancial inclusion. However, the potentially disruptive growth of rms oering novel products and services poses new challenges for nancial stability and consumer protection. 2 Policymakers around the world are stepping up their eorts to foster innovation in the nancial sector while keeping alert to emerging risks. A landmark initiative was the creation of the regulatory sandbox by the United Kingdoms Financial Conduct Authority (FCA) in November 2015. Sandboxes oer ntechs a controlled testing environment in which they can try out their products on a limited set of customers under restricted authorisation. Testing occurs under close regulatory supervision: rms receive advice to help them navigate the complexities of regulations and to ease the route to authorization. Regulators, on the other hand, use sandboxes to learn about new nancial technologies and emerging trends, as well as to identify associated risks before products are launched for the mass market. A key objective of sandboxes is to foster innovation by facilitating ntechs access to nancing at early stages of development. Since ntechs oer new products in an environment of high regulatory uncertainty, they face serious challenges of asymmetric information and often struggle to raise enough capital to develop products and expand. 3 By now, around 50 countries have followed the UK and introduced their own regulatory sandbox, often with the goal of nurturing the ntech sector (Wechsler, Perlman and Gurung, 2018; Schizas, McKain, Zhang, Garvey, Ganbold, Hussain, Kumar, Huang, 1 The Financial Stability Board denes the term ntech as: technologically enabled nancial inno- vation that could result in new business models, applications, processes, or products, with an associated material eect on nancial markets and institutions and the provision of nancial services (Financial Stability Board, 2017). 2 Fintechs often rely on sweeping technological advancements (such as articial intelligence, machine learning, blockchain technology, big data analytics, or the internet of things) that pose signicant pri- vacy, regulatory, and law-enforcement challenges. A further risk associated with ntechs is cyber risk (Aldasoro, Gambacorta, Giudici and Leach, 2020). 3 While investors enthusiasm for ntech start-ups has boomed since 2010, reaching over $200 billion across 5,000 deals worldwide in 2019, investments have been volatile, displaying for example a sharp decline in 2016 and 2017 (see Figure 1). 2Wang and Yerolemou, 2019). 4 And yet, despite the wide-spread adoption of sandboxes and signicant attention in the media and policy circles, little empirical evidence exists on whether sandboxes actually help ntechs raise funding. Nor is there any evidence on the underlying channels that could be at work. In this paper, we analyze how entering the FCAs regulatory sandbox aects ntechs ability to raise funding. We collect unique data on capital raised by ntechs in the UK for the period from 2014q1 to 2019q2. Our sample covers ntechs that joined the sandbox (treated rms), as well as a large group of comparable control rms. Granular data on funding raised, broken down by individual investor, as well as background information on rm age, size, industry, location, and CEO background allow us to investigate dierent channels through which the sandbox aects rms access to capital. Our main nding is that entry into the sandbox is associated with a higher probability of raising funding and an increase in the average amount of funding raised by around 15% (or $700,000), relative to rms that did not enter the sandbox. Investigating the mechanism, our evidence suggests that regulatory sandboxes reduce information asymmetries and regulatory costs or uncertainty. For identication, we rely on two complementary approaches. First, we focus on the sample of rms that are accepted into the sandbox and exploit the fact that these rms entered the sandbox in ve dierent cohorts. Entry is staggered over rounds of six months, allowing us to compare a rms capital-raising activity before and after participation in the sandbox, relative to rms that will enter the sandbox at a later stage. We nd a highly signicant and economically meaningful eect of entry on capital raised. Relative to rms that will enter the sandbox at a later date, entry into the sandbox is followed by a 14% to 15% increase in capital raised over the following two years. The increase in capital raised corresponds to about one standard deviation. Selection into the sandbox is not random as we discuss in Section 2 and the entry date could be correlated with unobservable rm characteristics. Yet, we show that there 4 At the international level, national regulators take part in the Global Financial Innovation Network, a global sandbox initiative led by the UKs FCA (Ehrentraud, Ocampo, Garzoni and Piccolo, 2020). See also a recent survey by the World Bank and Cambridge Center for Alternative Finance (CCAF) (2019) on regulating alternative nance. 3are no dierential pre-trends across rms, and that among the group of rms that enter the sandbox at some point, the specic entry date is uncorrelated with observable rm characteristics. Likewise, our results are robust to controlling for rm age, CEO gender, or location; and to the inclusion of rm xed eects. These facts mitigate concerns that our results are explained by omitted variables or selection eects. We also nd that including time-varying xed eects at the industry level does not aect the size or signicance of our coecients in a substantive way, despite more than doubling the R 2 . In other words, sandbox entry is likely orthogonal to unobservable time-varying industry characteristics, further reducing potential concerns about self-selection and omitted variable bias (Altonji, Elder and Taber, 2005; Oster, 2019). To further strengthen identication, in a second step we compare sandbox ntechs to a set of control rms that never enters the sandbox. Using a coarsened exact matching approach, we select a sample of matched control rms that are statistically similar in terms of observable rm characteristics: age, CEO gender, industry, and location. We then estimate a dierence-in-dierences specication with rm and time xed eects, comparing rms that enter the sandbox to those that never enter the sandbox. In the matched sample, we nd almost identical eects to our baseline strategy: entry into the sandbox is associated with a relative 15.1% increase in funding raised. After establishing that sandbox entry improves rms access to funding, we inves- tigate the underlying mechanisms. Specically, we distinguish between the following channels: rst, the sandbox is a marketing device, i.e. simply entering the sandbox leads to publicity and hence more funding, irrespective of actual rm performance or support. Second, the sandbox serves as a stamp of approval, i.e. it reduces informa- tion asymmetries, as being accepted into the sandbox signals high quality. And third, the sandbox reduces regulatory uncertainty or costs, i.e. the dedicated case ocer helps sandbox rms in navigating uncertainties about legal challenges to their services or products. Our results suggest that the sandbox reduces information asymmetries and reg- ulatory costs. We nd no support for the notion that sandboxes serve purely as a 4marketing device. 5 First, we show that the positive eect of sandbox entry on capital raised is particularly pronounced for smaller and younger rms, i.e. rms that are usu- ally considered more opaque and hence subject to severe informational frictions (Hall and Lerner, 2010). We nd similar results when we compare rms by type of fund- ing. Entry into the sandbox increases deal volume especially for venture capital deals, which are more information-sensitive, compared to other types of deals. Second, we show that entry into the sandbox is followed by an increase in rst-time investors and in the share of investors that are based outside the UK. Since new investors and investors that are located further away from the issuing rm are likely to face higher information asymmetries, we interpret this nding as evidence that the sandbox reduces information asymmetries. Finally, we show that rms with a CEO who has a personal background in (nancial) law benet less from entry into the sandbox. This is in line with anecdotal evidence that CEOs without prior experience in nancial regulation benet more from the guidance provided by case ocers (Deloitte, 2019). If sandboxes promote all rms through marketing irrespective of their underlying features then we should not nd any dierential eects across rm types. In principle, investors could learn about rms as their quality is gradually revealed to the market over time, irrespective of entry into the sandbox. Firms ability to raise funding would then increase in lockstep. Instead, if investors learn about the quality of a rm because of the sandbox certication, rms ability to raise funding will increase immediately upon entry. We nd that the strongest eects on funding raised occur in the rst two quarters upon entry. Four quarters after entry, the sandbox has a modest positive, but insignicant, eect on funding raised. This pattern hence suggests that entry into the sandbox acts as a certicate and signals rms quality. The increase in funding raised does not reect a gradual revelation of rms quality. We provide a set of further robustness checks. We rule out that the eect of the sandbox is driven purely by an increase in the supply of funds. We use matched investor- 5 In principle, a marketing eect could also disproportionately aect smaller or younger rms by raising awareness among investors. As long as the sandbox selects innovative rms, the marketing eect and stamp of approval eect would be complementary. For expositional clarity we decide to treat both eects as distinct. 5rm data and include investor*time xed eects. This approach allows us to absorb any unobservable time-varying changes aecting investors. For example, contemporaneous tax reliefs or preferential treatment for ntech investments could act as confounding factors. We show that a rm entering the sandbox is more likely to raise capital even if we hold changes in the supply of capital xed. We also use alternative estimation methods to account for the presence of zeros in our dependent variable, for example negative binomial regressions, and show that our results are insensitive to the chosen method. Further, we conrm our ndings when we use nearest neighbor matching or propensity score matching instead of coarsened exact matching. Despite the widespread adoption of sandboxes, to the best of our knowledge there exists no micro-evidence on their eectiveness. Regulatory sandboxes pursue dierent goals, for example promoting innovation and competition, increasing the consumer sur- plus, and facilitating ntechs access to nance. While the short time span since their inception does not allow us to evaluate eects on consumer surplus or nancial stability, our paper provides rst evidence that sandboxes help young and innovative ntechs to raise capital and hence achieve at least one of their explicit goals. Our results suggest that sandboxes could become a crucial policy tool for harvesting the benets of nancial innovation. Our paper contributes to the current debate on public policies to foster innovation (OECD, 2017; Auer, 2019). 6 A recent literature has established that ntechs face serious obstacles to raising capital (Block, Colombo, Cumming and Vismara, 2018; Haddad and Hornuf, 2019), despite the fact that their innovation provides value to innovators and investors (Chen, Wu and Yang, 2019). Market failures can lead to sub-optimal private- sector expenditure on research and development, necessitating public policies to foster innovation, eg through business incubators or accelerators (Gonzalzez-Uribe and Reyes, 2020). 7 Polic
展开阅读全文