我國金融科技創新實驗制度之運作成效檢討

金融監理沙盒是為金融科技蓬勃發展下為使其相容於現行法制下的產物，其自2016年英國提出以來便相繼受到多國採行，研究文獻亦有大量的產出：而迄今，對於斯的研究重點已由「是否應當設置監理沙盒」慢慢轉移至「制度上應如何設計為佳」；尤其現今主要採行沙盒的國家已累計一定的實驗經驗與案件量，俾由從其中進行分析以探討不同制度設計的優劣。　　本文乃先行介紹監理沙盒的起源脈絡－金融科技的發展，並點出各國監理者對其所採取之措施，而監理沙盒被普遍認為屬最具制度化且彈性化的措施，但亦非完美，概括上監理沙盒仍有過於偏向鼓勵創新而忽略風險控制、試驗中監理成本集中於金融業者一方、耗費監理資源、篩選標準不明確或無法操作、試驗誤差、造成認知捕捉及造成市場競爭不平等問題，但本文認為部分可能為金融創新面臨的三難困境中折衝下所生的短處，不宜因其而全盤否定監理沙盒的效果。而後本文介紹我國監理沙盒制度與實施情形：就後者而言截至2022年4月時，我國共計有9件沙盒案件，乃分別就業者背景、產品類別、豁免法規類別與後續落地修法情形等進行分析整理。其後，本文進行沙盒之比較法上分析，其中發現澳洲所實行的沙盒較主流國家者有所差異，不論在試驗主客體或核准機制均較不具可預見性及彈性空間，而亦反映於其低使用數之上。英國與新加坡於制度設計上較為相近，但於消費者保護機制及資訊透明度方面仍略有出入。　　本文文末乃嘗試就我國現行的運作情形分別以不同之指標進行分析，惟量化指標上因相關數據取得之困難而成效有限；最後，藉由上述分析得出對於我國監理沙盒制度設計上的建議以作為結論，分別為：由同業公會辦理豁免自律規範的試驗、屏除案件申請之「創新性」要件、試驗客體不宜僵化限制、監理沙盒的主管機關宜具有統合跨部會之能力、宜引入公務員免責條款、引進限制性執照等。

關鍵字

金融監理沙盒；金融科技；金融創新；創新實驗；法規豁免；業務試辦；法令鬆綁

並列摘要

The issuance of a regulatory sandbox is accompanied with the booming development of FinTech. Since its guide has been first released in the UK in 2016, many countries have followed suit soon after, and many related studies have also yielded. Until now, the focus of it has shifted from “the necessity of a sandbox” to “how to design useful mechanisms”. Especially the main developing countries have accumulated enough experimental experience and cases to analyze the pros and cons between different sandbox mechanisms. 　　This dissertation begins with the origin of a regulatory sandbox – the development of FinTech, and points out the coping strategies taken by each regulator and a regulatory sandbox has been considered as the most systematic and flexible. However, it’s also imperfect. In general, a sandbox has the following faults: adopting to further the goals to promote innovation rather than risk control, allocating most of experimental cost to financial institutions, much more regulatory capacity consumption, implicit selection criteria, experiment bias, cognitive capture, and unfair market competition. The dissertation argues that some of the shortcomings arising are the trade-off between the the innovation trilemma, and shouldn’t deny the effect of a sandbox. And then the dissertation introduces the domestic sandbox and experiment cases: regarding the latter, nine experiments cases were accepted until April, 2022, and then the dissertation analyzes and sorts the applicants’ background, products, exemption, law amendment after experiments and so on. After that, the dissertation does comparative law research and discovers that Australian sandbox mechanisms differ from those of the mainstream sandboxes: experimental subject and object or application process is less foreseeable and flexible. It also reflects in the lower amount of the applicants. The UK and Singaporean sandbox mechanisms are similar, but there are still differences in consumer protection and transparency of information. 　　In the end, the dissertation try to analysize the performance of the domestic sandbox in different kinds of indexes and then propose the solutions as the conclusion of the dissertation, however, there are not significant results in quantity indexes due to the limited data aquired. The solutions include: guilds can conduct experiments on the self-regulation, eliminate the eligibility criteria of “innovation”, remove the limit of experimental object, the competent authority should have the ability and statutory power to integrate the opinion from different departments, introduce the waiver to the civil servants, introduce the restricted licences and so on.