1、STATE OF SUPTECH REPORT 2022SUPPORTED BYThe Cambridge SupTech Lab at the Cambridge Centre for Alternative Finance,the University of Cambridge Judge Business School,acceler-ates the digital transformation of financial supervision.While financial services are becoming increasingly global,digital and c

2、omplex,analogue processing and antiquated technologies in data gathering,validation,storage and analysis erode the analytical capa-bilities of supervisory agencies,who are often too late in protecting consumers from fraud and seeing signs of stress in the financial sys-tem or miss the underlying cau

3、ses.This is all happening while financial crime remains a trillion-dollar issue,and public agencies face new challenges such as the regulation and supervision of crypto assets,and monitoring environmental,social and governance(ESG)aspects of the financial industrys business.The Lab aims to meet fina

4、ncial sector supervisors needs by working with them to develop new methodologies and processes that further market oversight and empower consumers,and to deployh suptech applications that generate relevant,reliable,timely insights to inform their decisions.From research to executive education,to tec

5、hnical assistance,to crafting production-grade suptech solutions,we are committed to supporting the emergence of the suptech ecosystem and to em-powering a new generation of innovation leaders seeking to digitally transform financial supervision.We invite you to find out more atABOUT THECamSupTechLa

6、bCambridesuptechlabwww.cambridgesuptechlab.orgAuthors Simone di CastriMatt GrasserJuliet OngwaeJose Miguel MestanzaDesignMarta LoperaEmily DuongJessica AliCopy editingAlpa SomaiyaAdebola Daramola,Alexander Apostolides,Kyriakos Christofi,Philip Rowan,Yue Wu and Bryan Zhang of the Cambridge Centre for

7、 Alternative Finance(CCAF)contributed to the analysis.Suggested citation:Cambridge SupTech Lab(2022),State of SupTech Report 2022,Cambridge:Cambridge Centre for Alternative Finance(CCAF),University of Cambridge.Available atwww.cambridgesuptechlab.org/SOSThe mention of specific companies,manufacturer

8、s or software does not imply that they are endorsed or recommended by the Cambridge SupTech Lab in preference to others of a similar nature that are not mentioned.All graphics and charts can be downloaded at www.cambridgesuptechlab.org/SOSThe Cambridge SupTech Lab is supported byCAMBRIDGE SUPTECH LA

9、B TABLE OF CONTENTSEXECUTIVE SUMMARY.6SAMPLE,METHODOLOGY,AND TAXONOMY.101.1.Research methods.111.1.1.Sample of financial authorities by geography and income classification.111.1.2.Questionnaire for financial authorities on specifics of supervisory data.131.1.3.Questionnaire for suptech vendors.131.2

10、 Suptech taxonomy.151.2.1.Supervisory areas and use cases.151.2.2.Technologies and data science tools in the supervisory stack.16EVOLUTION OF THE SUPTECH LANDSCAPE.182.1.Timeline of the digital transformation of financial supervision.192.1.1.19872007:Suptech foundations.212.1.2.20082016:The global f

11、inancial crisis and the mass adoption of fintech.212.1.3.20172019:The dawn of suptech.222.1.4.2020present:Covid-19 accelerates suptech.23THE STATE OF SUPTECH .243.1.Demand:Financial authorities .253.1.1.Adoption .253.1.2.Gaps .303.1.3.Suptech generations 2.0.333.1.3.1.Data collection .363.1.3.2.Data

12、 processing.373.1.3.3.Data storage.373.1.3.4.Data analytics.383.1.3.5.Data products.383.1.4.Supervisory areas.393.1.5.Enabling factors.403.1.6.Funding.403.1.7.Governance.423.1.8.Gender.463.1.9.Outcomes.483.2.Supply:Sourcing solutions.503.2.1.Sources of suptech apps.503.2.2.The vendors business case.

13、513.2.3.Offerings by focus area.523.2.4.Funding.524|STATE OF SUPTECH REPORT 2022CHALLENGES TO UPTAKE.534.1.Challenges:financial authorities.544.1.1.Implementation.544.1.2.Data lifecycle.574.1.3.Resources.574.1.4.Infrastructure.614.2.Challenges:vendors.61CASE STUDIES.645.1.Data collection:Bank of Eng

14、land transforming data collection from the UK financial sector.655.2.Data processing:Central Bank of the Philippines API-based prudential reporting system and back-office reporting and visualisation application.695.3.Data storage:National Bank of Rwanda Electronic Data Warehouse.735.4.Data analytics

15、:Central Bank of the Netherlands outlier detection tool for AML/CFT/PF supervision.765.5.Data products:Reserve Bank of India(RBI)DAKSH.815.6.Full stack:BIS Project Ellipse,an integrated regulatory reporting and data analytics platform.82CONCLUSIONS.86Develop a suptech strategy and/or roadmap.87Build

16、 data capabilities for the supervisors of the future.88Grow a data-driven innovation culture.88Scale.89References.90Appendix 1:List of respondents .95Appendix 2:Suptech Taxonomy.102Appendix 3:Definitions.106CAMBRIDGE SUPTECH LAB EXECUTIVE SUMMARY6|STATE OF SUPTECH REPORT 2022The State of SupTech Rep

17、ort 2022 focuses at how financial authorities are developing and implementing supervisory technologies(suptech),and establishes a baseline from which to track the progress and impact of suptech adoption allowing financial authorities across the world to benchmark the progress of their suptech initia

18、tives.To facilitate more granular analyses of these macro trends,the Report introduces a novel version of the“SupTech Taxonomy”adopted by the Bank for International Settlements(BIS)(BIS 2018,BIS 2019),classifying supervisory use cases,technologies,and data science tools in a standardized and structu

19、red manner.In order to complement the analyses and to ground the findings in a practical context,the Report also provides a timeline of disruptions and innovations in supervision,and a set of six case studies of suptech applications.The Report is based on the insights that 146 financial authorities

20、shared through:A survey of 134 financial authorities from 108 jurisdictions A questionnaire on data models with 74 individual supervisors representing 46 agencies and 35 jurisdictions.The analysis also advances the understanding of the suptech marketplace from the supply side,providing critical insi

21、ghts from the nascent but rapidly growing industry of suptech vendors through in-depth qualitative research of key vendors sampled from the Cambridge SupTech Labs SupTech Marketplace and highlighting their perspectives on the business case for suptech,the primary use cases they focus on and the chal

22、lenges they face in commercializing suptech solutions.The Cambridge SupTech Lab State of SupTech Report 2022 presents insights on the current state of the digital transformation of financial supervision worldwide.The Report provides a global snapshot across several facets of suptech,including underp

23、inning digital infrastructure and technologies,supported supervisory use cases,approaches employed for developing and deploying suptech applications,and the related challenges and risks.CAMBRIDGE SUPTECH LAB|7 Suptech is happening.Most financial authorities have already engaged in suptech initiative

24、s.While suptech development is still at a nascent stage with room for growth,the survey results indicate that 71%of financial authorities are rising to the challenge as we see the adoption of suptech solutions,strategies and roadmaps increasing.Suptech efforts remain in the experimentation stage,pri

25、marily focused on improving data collection and basic analysis.Based on the classification provided by the Bank for International Settlements(BIS 2019)and revised by the Lab in this report(see chapter 3),the technologies deployed by financial supervisors mostly fall into the first or second generati

26、on of data architecture,and mainly support data collection as well as descriptive and diagnostic analytics.Most suptech use cases centre around consumer protection and prudential supervision.59%of financial authorities report their suptech applications being deployed in support of consumer protectio

27、n supervision,while 58%report their suptech applications support prudential supervision use cases.Significant challenges to suptech adoption remain to be addressed.Limitations in budget,data quality and technical skills remain the most significant barriers to implementing suptech.There is a remarkab

28、le mismatch between the experience of financial authorities and vendors when it comes to procurement,with technologies providers urging the public agencies to address legacy procurement processes.Financial authorities also express an unmet need for data teams,data sharing and data synthesis as a fou

29、ndational part of their modernization.There are significant distinctions in the state of suptech in emerging markets and developing economies(EMDEs)as compared to advanced economies(AEs).Financial authorities in AEs are early adopters of suptech,more often have sufficient digital infrastructure,more

30、 often assign dedicated suptech roles and departments,have seen more substantial internal outcomes than those in EMDEs,and seek funding primarily to grow their teams.EMDEs agecies tend to run suptech initiatives within the supervision department itself,are more interested in trainings,technical assi

31、stance,digital tools,and seek funding primarily for design and development of suptech.Financial authorities in EMDEs and in AEs face very similar challenges in the digital transformation of their supervisory process and capabilities.Agencies in EMDEs and AEs report lack of budget being the main cons

32、traint to the development and deployment of suptech.Centralised data office models to accelerate suptech development and implementation are emerging.35%of the surveyed financial authorities have a dedicated centralised office reporting to a Chief Data Officer who is either solely responsible for the

33、 suptech initiatives or works with other functions to develop and deploy suptech.Funding to accelerate the suptech market is a key area of focus.Although suptech vendors report some secondary support from grants,funding for financial authorities suptech initiatives comes primarily from the financial

34、 authorities themselves.Most suptech solutions are provided by external sources Highlights FROM THE state of suptech REPORT 20228|STATE OF SUPTECH REPORT 2022like contracted vendors and purchased off-the-shelf software,yet these vendors also report challenges in funding and an ability to deeply unde

35、rstand financial authorities prioritized needs.The top suptech challenges differ between agency types.For central banks,the challenges are primarily related to internal culture and strategic buy-in.For capital markets,securities,and investment instruments supervisors,challenges tend to be related to

36、 upgrading their existing systems and processes.For other supervisors,the uniquely prominent challenges are with IT systems.Most authorities still do not have a gender data strategy.Only 21%have a currently operating strategy,9%have one in development,while 70%report no strategy at all.Suptech is en

37、abling new supervisory use cases that would not otherwise be possible.While suptech solutions use chatbots and APIs to optimize existing processes and augment legacy tools,others are opening completely new opportunities for supervisors.The ability to ingest massive online datasets like social media

38、streams to conduct sentiment analysis,to parse online reviews to assess risks or identify fraudulent fintech apps,and to conduct real-time,on-chain analyses for digital assets supervision are just a few of many examples.Taken on the whole,these insights frame a suptech space that is relatively nasce

39、nt,but rapidly and necessarily accelerating to address the needs of supervisors in the face of novel and newly-magnified risks introduced by a financial sector that is digitalizing and generating supervisory data at an exponential rate.Addressing the needs of the ecosystem in an effective and equita

40、ble manner will require close collaboration between financial authorities,vendors,funders,educators,researchers,technologists,data scientists,and the rest of the suptech ecosystem.This inaugural annual State of SupTech Report aims to feed that conversation and support collaboration,building a baseli

41、ne against which to conduct agency and regional benchmarking,methodically tracking year-on-year trends,and a growth of a marketplace to serve the needs of supervisors,who in turn serve the interests of the billions of financial citizens of the jurisdictions they oversee.CAMBRIDGE SUPTECH LAB|9SAMPLE

42、,METHODOLOGY,AND TAXONOMY1.10|STATE OF SUPTECH REPORT 2022Three primary data sources were used to compile this report:A survey of 134 financial authorities from 108 jurisdictions A questionnaire for 74 individual supervisors(representing 46 agencies and 35 jurisdictions)on the specifics of superviso

43、ry data A questionnaire for six selected suptech vendors.In addition,the Lab complemented these resources with qualitative interviews and case studies to further develop and test hypotheses arising from the quantitative data and more deeply understand the challenges and opportunities in adopting sup

44、tech applications.Most of the data presented in this Report were collected between May and October 2022 through a global survey conducted by Cambridge SupTech Lab.The respondents include financial authorities such as central banks,securities and capital market authorities,financial conduct authoriti

45、es,and insurance regulators.Of the 134 responses,81 are from central banks,representing 60%of the total sample.92 responses were received from agencies in emerging markets and developing economies(EMDEs),representing 67%of the responses,while the remainder were from advanced economies(AEs).1.1.Resea

46、rch methods1.1.1.Sample of financial author-ities by geography and income classificationFigure 1.Geographical distribution of survey respondentsNumber of Agencies PER COUTRY123CAMBRIDGE SUPTECH LAB|11TABLE 1.Geographical distribution of respondents By regionRegionNumber of respondentsPercentage of s

47、ample by regionPercentage of jurisdictions covered within region East Asia and the Pacific2216%46%Europe and Central Asia2922%41%Latin America and theCaribbean2720%44%The Middle East andNorth Africa1410%46%North America32%100%South Asia65%63%Sub-Saharan Africa3325%48%Total134*Income and region are b

48、ased on the World Bank Country Classification.If a jurisdiction was not listed geo-graphically,its classification was based on neighboring jurisdictions.Figure 2.Breakdown of respondents by income groups(N=134)The final respondent sample is geographically diverse and representative of World Bank Cou

49、ntry income groups.Table 1 maps the 108 geographic jurisdictions of the 134 financial authorities who responded to the survey.The complete list is available in Appendix 1.Figure 2 illustrates the response distribution according to the World Banks classification by income level.The sample contains re

50、sponses from jurisdictions across all four income classifications,with 55 responses from either low or lower-middle-income jurisdictions.In some areas of the analysis,we group these categories into EMDEs(low,lower-middle and upper-middle income)and AEs(high income).12|STATE OF SUPTECH REPORT 2022eIn

51、 November 2022,we asked individual supervisors four questions on the specifics of supervisory data to further assess the state of data collection for financial supervision:1.Thematic areas:the supervisory areas for which data is collected2.Channels:the mechanisms and channels through which it is col

52、lected3.Formats:the digital format and structure of data that is collected4.Challenges:the specific challenges faced at each layer of the supervisory data lifecycle stackWe received information from 74 supervisors representing 46 agencies and 35 jurisdictions.This sample included some supervisors wh

53、ose agencies did not participate in the primary survey,whose agencies are listed in Appendix 1.1.1.3.Questionnaire for suptech vendorsTo complement the insights shared by the demand side of the suptech market and develop a deeper understanding of the broader suptech ecosystem,we also engaged directl

54、y with six suptech vendors to discuss ten questions that characterise the opportunities,challenges and other qualitative characteristics of the market.The vendors were selected from the Cambridge SupTech Labs SupTech Marketplace Vendor Database based on the following criteria:Centricity of suptech i

55、n strategic focus:While some vendors provide suptech solutions as a small part of a broader portfolio of products and services,others focus primarily on suptech solutions.For this set of interviews,we prioritised the latter.Maturity of offering:The sample prioritised vendors with a mature product or

56、 service to ensure actual experiences inform interviews of operating in the market,not hypothetical or early-stage ideas based only on pilots or experiments.Diversity of market position:The sample aimed to incorporate a range of market perspectives,including relatively new entrants(those who have on

57、ly recently adapted their mature offering to address supervisory use cases)and those who have been working with supervisors since before the inception of the word suptech.Diversity of geographies where solutions are deployed:The sample aimed to capture experiences across a range of jurisdictions to

58、avoid sample bias toward any one set of cultural norms or localised market restrictions.1.1.2.Questionnaire for financial authorities on specifics of supervisory dataCAMBRIDGE SUPTECH LAB|13Figure 3:Suptech taxonomy14|STATE OF SUPTECH REPORT 2022The 13 thematic focus areas are:Anti-Money Laundering/

59、Countering the Financing of Terrorism/Financing the Proliferation of Weapons of Mass Destruction(AML/CFT/PF)supervision:Suptech allows financial authorities to identify potentially suspicious customers or activities(for example,through customer due diligence and suspicious transactions detection)and

60、 enhances data analytics to monitor institutions compliance and AML/CFT/PF risk management(for example,assisted/automated examination,metadata analytics,and text analytics).Capital markets,securities and investments supervision:Suptech equips financial authorities to detect potential misconduct(for

61、example,insider trading,market manipulation and poor disclosure)and enhances data analytics to monitor the capital markets(for example,automated examination,peer-group/risk classification and text analytics).Securities and investments use cases focus on empowering securities commissions and other fi

62、nancial authorities with a securities mandate to augment their capabilities by generating improved data-driven insights and detecting insider trading and market manipulation.Climate/ESG risk supervision:Suptech enables financial authorities to enhance data collection and analytics to assess institut

63、ions climate and environment,social and governance(ESG)risk management(for example,green market monitoring,peer-group/risk classification and stress testing).Competition monitoring:Suptech focuses on monitoring market competition dynamics and rates and fees.Compliance assistance:Suptech makes availa

64、ble automating compliance auditing and automated guidance for compliance queries.The Cambridge SupTech Lab has developed a comprehensive classification system to consistently organise various entities namely,suptech vendors,suptech solutions and suptech diagnostics by supervisory use case(the sup in

65、 suptech)and by the technologies and data science tools used(the tech).This taxonomy is based on past efforts to map the space(BIS 2018,BIS 2019)and explicitly differentiates between the sup and the tech.This disaggregation affords a novel opportunity to systematically map the needs of supervisors,c

66、lassify the tools serving those needs and ultimately serve as an ontology for connecting the solutions to needs strategically and intentionally.It was refined and validated through desk research,review of deployed suptech applications(see the Labs SupTech Marketplace),and input from over 130 financi

67、al supervisors and leading suptech experts.The taxonomy will be periodically revised,based on internal research and external feedback,to reflect the suptech spaces dynamic nature.1.2.1.Supervisory areas and use casesThis first iteration of the taxonomy covers 13 broad supervisory categories subdivid

68、ed into 87 use cases.The structure of the classification system is hierarchical and built on a conceptual framework that groups use cases according to the activities conducted by supervisory functions within authorities.While thematic focus areas refer to policy or supervisory areas/activities,use c

69、ases refer to more specific tasks supported by identified suptech tools.1.2 Suptech taxonomyCAMBRIDGE SUPTECH LAB|15Consumer protection and market conduct supervision(now referred to as consumer protection):Suptech empowers financial authorities to enhance data collection(for example,advanced/real-t

70、ime monitoring and data consolidation)and improve data analytics to monitor consumer risks and supervise market conduct(for example,assisted/automated examination,misconduct detection,peer-group/risk classification and text analytics).In addition,these use cases also support authorities in providing

71、 consumers with virtual assistance(for example,complaints handling and credit bureau rectification).Cyber risk supervision:Suptech improves data analytics to monitor institutions compliance and cyber risk management(for example,automated examination,assessment of vulnerabilities and compliance monit

72、oring).Digital assets supervision:Suptech is deployed to supervise cryptoassets or DLT-based protocols,platforms or systems(for example,cross-jurisdictional intelligence checks and information-sharing capacity,embedded supervision and on-chain analysis).Financial inclusion:Suptech is used by financi

73、al authorities to monitor the access and use of financial services(for example,gender-based and geospatial analysis).These use cases can also collect consumer data(for example,consumer satisfaction analysis)and provide virtual assistance(for example,financial education tools).Insurance supervision:S

74、uptech serves some prudential supervision use cases that enable insurance supervisors to enhance data collection and data analytics.In addition,and covers use cases that allow insurance supervisors to provide virtual assistance to firms for procedures often required in the insurance sector(for examp

75、le,registration of intermediaries and product registration).Licensing:Suptech supports financial authorities providing virtual assistance to firms requesting a license or authorisation to operate within the regulatory perimeter(for example,automated guidance and automated processing of requests).Pay

76、ments oversight:Suptech assists financial authorities in monitoring and testing the performance of payments infrastructures,networks and systems(for example,advanced/real-time monitoring and stress testing).Prudential supervision of banks and non-bank deposit-taking institutions:(now referred to as

77、prudential supervision):allows financial authorities to enhance data collection(for example,automated reporting,automated validation and data consolidation)and data analytics for both macroprudential and microprudential supervision(for example,assisted/automated examination,peer-group/risk classific

78、ation and stress testing).The complete list of suptech use cases grouped by thematic focus area is available in Appendix 2.1.2.2.Technologies and data science tools in the supervisory stackOn the other side of the taxonomy in Figure 3 are the technologies and data science tools deployed to address a

79、uthorities challenges and realise the aspirations within the aforementioned supervisory areas and use cases.These technologies are classified by their applications within the context of the five layers of a supervisory stack(RA 2020):Data collection:This is the layer where data is gathered.It is col

80、lected from entities,including supervised financial service providers,consumers of financial technologies and the general public,into the supervisors domain.Examples of data collection mechanisms used for supervision include web portals and other 16|STATE OF SUPTECH REPORT 2022document management,ap

81、plication programming interfaces(APIs),advanced collection techniques including AI-based tools like chatbots,embedded supervision of distributed ledger technologies(DLT),and automated data gathering like web scraping and data streaming.Data processing:As the data is being gathered,it should be valid

82、ated,cleaned and consolidated using data processing tools to maximise its utility.Examples of data processing technologies in the supervisory context include integrated validation techniques like rules on APIs that send errors back to the submitting party in real time,task automation techniques like

83、 those programmed in scripting languages or recorded and replayed via robotic process automation(RPA),and advanced processing tools such as machine learning based computer vision and natural language processing models to extract structured supervisory data from less structured sources.Data storage:O

84、nce the data has been collected and processed,it needs to be stored in a manner that ensures security and ease of access across supervisory areas.Examples of storage methods for supervisory data include databases hosted and managed onsite by the financial authority itself,cloud and hybrid computing

85、technologies that introduce the benefits of virtualisation,and big data tools such as data lakes and data warehouses.Data analytics:With the data suitably stored,extracting insights can begin,a process enabled by data analytics technologies.Examples of data analytics tools used by supervisors includ

86、e descriptive and diagnostic analytics that summarise the current moment in time,predictive analytics that create statistical models from historical data to infer the most likely outcome in the future,and prescriptive analytics tools that use those predictions to recommend the most effective action

87、the financial authority can take to optimise achieving their mandates and goals.Data products:At the top of the stack are the products and interfaces that directly connect supervisors to the insights derived from the analytics.Examples of data products for financial authorities include charts and ke

88、y metrics from static reporting tools,interactive visualisations and dashboards that allow deeper exploration and combinations of data,and advanced business intelligence tools that leverage artificial intelligence(AI)to deliver alerts proactively.CAMBRIDGE SUPTECH LAB|172.EVOLUTION OF THE SUPTECH LA

89、NDSCAPE18|STATE OF SUPTECH REPORT 2022The use of technology and data science for financial supervision and market monitoring has rapidly evolved over the past two decades.2.1.Timeline of the digital transformation of financial supervisionIn part,this evolution has been a conjunctural phenomenon,a re

90、sponse to events-sometimes endogenous,other times exogenous to the financial system-that have reshaped financial supervision.Such events include international terrorism in the 2000s,major financial scandals at the beginning of the same decade and the global financial crisis in 2008,and more recently

91、 the Covid-19 pandemic.Moreover,this evolution also reflects a structural shift connected to the digitization of the financial market and the exploitation of big data by financial firms,driven by progress in technology and computing power,and their increased availability and affordability.Along with

92、 this progress comes the introduction and magnification of risks,such as cybersecurity and data privacy,which become ever more prominent with the advent of this digital era and proliferation of abundant digital financial data.In this context,financial authorities have increasingly experienced a digi

93、tal flood of supervisory data,without being able to distill more intelligence to govern the financial sector.Therefore,supervisory agencies have started to re-engineer their institutional arrangements,rescope their mandates,review their risk management frameworks,readjust their methodologies,step up

94、 their data management and governance approaches,and enhance their competencies and capabilities to further their digital transformation.Notably,the suptech era appears to be only the most recent chapter in the broader anthology of tech-enabled innovation in financial supervision.This section frames

95、 suptech in that broader context,highlighting some of the key milestones along the ongoing journey toward a suptech-augmented,responsible,and resilient approach to financial supervision.CAMBRIDGE SUPTECH LAB|19Figure 4.Timeline for the evolution of suptech 1920172008 872020The dawn of suptechSUPTECH

96、 FOUNDATIONSCOVID-19 ACCELERATES SUPTECHCovid-19 pandemic,lockdowns,and physical distancingMass adoption of fintech product and servicesPost-crisis increase of reporting requirementsCreation of BitcoinGlobal financial crisisFinancial scandals and the Sarbanes-Oxley ActBlack Monday market crashAdopti

97、on of XBRL by supervisory agencies U.S.Securities and Exchange Commisssion adopts EDGAR Emergence of the blockchain technology Foundations of API,cloud computing,big data and AI/MLWeb-based portals&automation in the regulatory data pipelineFormal adoption of the term“SupTech”The Central Bank of the

98、Philippines launches chatbot to collect com-plaints from users,and deployes an API-based applications to collect regulatory reports The National Bank of Rwanda develops a data warehouse2020201719872008THE GLOBAL FINANCIAL CRISIS AND THE MASS ADOPTION OF FINTECH20|STATE OF SUPTECH REPORT 20222.1.1.19

99、872007:Suptech foundationsAfter the Black Monday market crash of 1987,regulators and supervisors began to digitise their operations to improve transparency and risk management in the financial markets.In 1993,the United States Securities and Exchange Commission mandated electronic filing through its

100、 Electronic Data Gathering,Analysis,and Retrieval(EDGAR)system to integrate digital disclosure requirements by supervised entities and enable monitoring of compliance and enforcement by supervisors(ADB 2022).In 1994,a risk management tool called Value at Risk(VAR)was developed to compute capital req

101、uirements and assess market risks per the Basel Framework.In the United States,the Sarbanes-Oxley Act,signed into law in 2002,marks an important moment because following the crash of Enron,Tyco,and WorldCom after two decades of deregulation the U.S.congress reaffirms the need for financial accountab

102、ility impacts every public company in the United States with a huge impact on the acceleration of solutions for the storage and management of corporate electronic records.Through the 1990s and 2000s,the financial authorities data management workflows were still heavily manual.Data collection relied

103、on paper forms,then emails and portals and Excel files.Data was stored in disjointed databases or folders.Validation occurred manually through spot checks,and the analysis primarily generated static management reports.In the second half of the 2000s,though,a wave of modernization began when standard

104、 data reporting formats started being incorporated into supervisory processes and technologies.The incremental adoption of the XBRL format by supervisory authorities on a global scale as reported by the United States Securities and Exchange Commission,the capital Market Authority in Chile(XBRL 2009a

105、),the Bank of Japan(XBRL 2009b),the Reserve Bank of India(XBRL 2009c)and others marked a critical moment in the suptech timeline.2.1.2.20082016:The global fi-nancial crisis and the mass adoption of fintechThe 2008 global financial crisis triggered a seismic shift in the financial landscape.The repor

106、ting and supervisory mechanisms were insufficient and did not allow financial authorities to detect the irresponsible actions of financial institutions through their predatory lending and subprime mortgage practices.As a result,many people were left financially exposed and dejected while taxpayers h

107、ad to bear the cost of bank bailouts.Authorities responded with stringent regulatory reforms that increased reporting requirements(BIS 2018).This encouraged the industry to develop new technologies to streamline regulatory reporting.On the other end,supervisory agencies increasingly automated their

108、procedures for data collection and management adopting web-based portals or bulk uploads to allow financial firms to file regulatory returns accompanied by inbuilt automated validation checks when uploading.The post 2008-crisis years earmarked another heavy increase in the number and complexity of t

109、emplates to be reported for prudential purposes,with banks required to report layers upon layers of harmonised reporting templates in digital formats(after the 2014 reforms,8,000 European banks were reporting up to 700,000 data points quarterly,while in 2018 HSBC announced that it was linking up 10

110、petabytes of data from over 300 data sources in a data lake,which is bigger than the entire Internet was at the turn of the century)and therefore making major investments in the development of regulatory technologies(regtech)solutions to handle compliance more efficiently.Groundbreaking approaches t

111、o applying advanced technologies to large financial datasets also emerged during this period,such as business intelligence through integrated management of micro-databases(BIS 2014),social media CAMBRIDGE SUPTECH LAB|21sentiment analysis to monitor consumer confidence(ECB 2014),and extensive set of

112、publications around the big data strategies for central banks as an explicit reaction to the“data revolution”(BOE 2015,Riksbank 2015).The resulting larger volume of collected data was translated into dynamic data visualisation through business intelligence dashboards and diagnostic analytics(for exa

113、mple,via scorecards),enabling richer insights.In addition,the mass adoption of fintech in the aftermath of the crisis pushed supervisory agencies to readjust their supervisory processes to keep pace with innovations in the financial sector.Mass adoption of cell phones in EMDEs and the online shoppin

114、g in AEs propelled the success of innovative products such as mobile money products in Africa(e.g.,mPesa)and mobile banking(e.g.,PayPal)in the United States.Advances in technology allowed start-ups and firms outside the traditional financial sector to develop disruptive business models,such as crowd

115、funding and peer-to-peer lending.Furthermore,decentralised finance,a new paradigm based on blockchain technology,emerged with the creation of bitcoin.The substantial rise in the volume and availability of data produced concerning digital and traditional markets prompted financial authorities to seek

116、 technological solutions that could support them2.1.3.20172019:The dawn of suptechWhile the use of innovative technologies for supervisory purposes accelerated throughout the 2010s,it was only in 2017 that the term suptech was introduced more formally into the conversation by Ravi Menon,Managing Dir

117、ector at the Monetary Authority of Singapore(MAS),to refer to supervisory technologies.In his view,technological innovation was necessary for financial authorities to reduce inefficiencies and make supervision more effective(Menon 2017).Like the MAS,many other authorities began to adopt an instituti

118、onal approach toward suptech,which became an object of interest to global standard-setting bodies.In 2017,the Basel Committee on Banking Supervision(BCBS)recommended that supervisors should consider exploring the potential of new technologies,such as artificial intelligence and machine learning,dist

119、ributed ledger technology,cloud computing and APIs,to improve their methods and processes(BCBS 2017).In 2018 and 2019,the BIS FSI published two seminal papers that provided an overview of the developments in the suptech ecosystem.The first report(BIS 2018)explored the experiences of ten early suptec

120、h users,highlighting the benefits and challenges and the implications for supervisors.They also proposed a taxonomy of areas of financial supervision in which suptech applications are used.This suptech taxonomy was slightly expanded in the second report(BIS 2019)who examined a sample of 99 suptech i

121、nitiatives and traced the evolution of the different generations of technology used by financial authorities,namely the suptech generations.A new generation of suptech applications went in production in this period,powered by application programming interface(API)and natural language processing,and

122、under the leadership of EMDEs financial authorities.In 2018,the National Bank of Rwanda deployed an electronic data warehouse that pulls data directly from the IT systems of 600 supervised financial institutions,including commercial banks,insurance companies,microfinance institutions,pension funds,f

123、orex bureaus,telecom operators and money transfer operators.Meanwhile,the RegTech for Regulators Accelerator(RA)partnered with the Bangko Sentral ng Pilipinas(BSP)and the Mexican Comisin Nacional Bancaria y de Valores(CNBV)and three technology firms to test and develop three suptech prototypes,namel

124、y an API-based prudential reporting system(RA 2018a),a chatbot application and consumer complaint management system(RA 2018b),and an AML data infrastructure and advanced 22|STATE OF SUPTECH REPORT 2022analytics solution(RA 2018c).2.1.4.2020present:Covid-19 accelerates suptechBuilding on the experime

125、nts of the previous decade,and consistently with the digital transformation of most industries and sectors of the economy,financial authorities are investing in the development of suptech applications.Some recent innovations are quickly scaling and being adopted in multiple countries-e.g.,chatbots f

126、or consumer protection have been depolyed in 2022 in Rwanda and Ghana while a number of other countries have put them in production,while APIs to collect and validate industrys data will soon become a standard.The budgetary section of the European Central Bank Annual Report on supervisory activities

127、(ECB 2021)is indicative of the increasing pace and magnitude of the suptech transformation(di Castri 2022):“The 7.9%increase in expenditure compared with 2020 mainly reflects the onboarding of new IT systems dedicated to banking supervision.With respect to the developments in IT systems,the increase

128、s in expenditure seen in the policy,advisory and regulatory functions 16.1%as well as macroprudential tasks 80.6%relate to the SSM information management system(IMAS)and the Stress Test Account Reporting platform(STAR).The main increase in expenditure in the Supervisory Board,secretariat and supervi

129、sory law section 43%resulted from significant investment in 2021 in supervisory technologies(suptech),which exploit the potential of artificial intelligence and other pioneering suptech in the context of banking supervision,for internal and external stakeholders.”Covid-19 has further accelerated thi

130、s modernization push.During the pandemic,many jurisdictions implemented measures such as lockdowns and physical distancing to reduce physical contact,which necessitated the urgent acceleration of digital financial services for the payment of goods and services,welfare transfers,etc.Digital transacti

131、ons increased dramatically and,unlike in 2008 when financial innovation was at the centre of the crisis,technology and finance were instrumental in responding to the crisis and supporting the recovery(ADB 2022).One of the most significant challenges for financial authorities during the pandemic was

132、adapting their processes to remote working.On-site supervisory activities and interactions with financial institutions and their staff had to move online.This made necessary the upgrade of the agencies infrastructure to allow for remote access to data bases,and accelerated the development of new sup

133、tech tools for qualitative scrutiny and risk assessments(BIS 2021).The use of natural language processing(NLP)helped automate the review of voluminous documents to identify corporate governance and credit risks.Additionally,analysis of responses to our 2022 survey shows that financial authorities pe

134、rception of increased consumer risk due to Covid-19 is correlated with their adoption and development of suptech applications.62%of those who had reported increased risks had an initiative or were developing an application,compared to just 44%of those who had not reported increased risks.A further 2

135、6%of respondents who identified increasing risks had taken no action to improve their supervisory capabilities using suptech versus 41%who did not identify increased consumer risk.This is in line with previous research findings that suptech adoption was accelerated by the Covid-19 pandemic,which nec

136、essitated a shift towards off-site supervision(CCAF and World Bank 2020,World Bank 2020)and an increased focus on using suptech applications to better address increased consumer vulnerabilities.CAMBRIDGE SUPTECH LAB|23THE STATE OF SUPTECH 3.24|STATE OF SUPTECH REPORT 2022THE STATE OF SUPTECH 3.1.Dem

137、and:Financial authorities Figure 5.Responses to the question Do you have any initiatives in your organisation that you would consider as suptech?(N=134)3.1.1.Adoption 71%of financial authorities have suptech initiatives.While suptech development is still at a nascent stage with room for growth,the s

138、urvey results indicate that financial authorities are rising to the challenge as we see increased adoption and use of suptech solutions.71%of the financial authorities surveyed indicated that they have already engaged in different suptech efforts.Three quarter of the respondents have one or more sup

139、tech applications in operation(50%)and/or an application in development(27%).A further 23%of the authorities reported that they have a suptech strategy or roadmap.Financial authorities in advanced economies are early adopters.While advances in suptech adoption are evident worldwide,financial authori

140、ties from AEs are early adopters of suptech applications.50%of financial authorities who stated they had already deployed suptech applications were from AEs,compared to 31%from EMDEs.CAMBRIDGE SUPTECH LAB|25Figure 6.Responses to the question Do you have any initiatives in your organisation that you

141、would consider as suptech?,segmented by income level(N=134)26|STATE OF SUPTECH REPORT 2022Securities supervisors lead in suptech adoption.Central banks lag behind.Securities industry supervisors take the lead in suptech adoption,while central banks lag behind.The reason might be that the former rely

142、 more on off-site supervision than central banks do,and up-to-date analytics software is needed to analyse massive volumes of transaction data.For example,the Australian Securities and Investments Commission(ASIC)has adopted suptech to transform data sets into usable patterns for market surveillance

143、 and suspicious trading detection.Securities supervisors core risks are of conduct nature rather than prudential.Often,these risks are evidenced in the processing of large unstructured information documents such as compliance manuals and internal policies,where new technologies allow supervisors to

144、extrapolate content from different firms to obtain an industry-level view that was previously just sensed by them.The three most adopted supervisory technologies are descriptive and diagnostic analytics,web portals and document management,and APIs.Investment in suptech applications primarily occurs

145、at the data-collection layer of the supervisory stack,followed by data analytics and data storage.CAMBRIDGE SUPTECH LAB|27Figure 7.Underpinning technologies used by agencies to enable supervisory processes(N=134)28|STATE OF SUPTECH REPORT 2022Figure 8.Underpinning technologies used by agencies to en

146、able supervisory processes,organised by layer of the supervisory stack of the suptech taxonomy(N=134)CAMBRIDGE SUPTECH LAB|293.1.2.Gaps Even with the adoption trends noted above,there is a substantial demand for improvement that arises from factors such as:Composition Each suptech application is but

147、 one component of a larger system,and as such the whole supervisory stack can be composed of suptech building blocks.In this sense,one suptech solution often begets the next.For example:A suptech data collection mechanism that processes higher volumes of data,then demands a storage mechanism within

148、which to store these data.In turn,a more advanced suptech data storage solution offers more accessible and robust datasets,which unlocks a stronger demand for analysis.EvolutionAn agency may choose to undergo digital transformation using less advanced technologies to soften the learning curve and re

149、duce the perceived costs of change management.Once the technology is adopted and a cultural shift to digital-first occurs,demand for more advanced technologies may arise.This evolution is reflected in more detail in the“suptech generations”section below.86%want prescriptive analytics,which use data

150、to guide them on what actions to take based on historical data.81%want task automation to record and replay tasks on a supervisors behalf to save time.79%want advanced image processing,such as computer vision in general,or more specific components like optical character recognition.78%want predictiv

151、e analytics,which can take past trends and forecast whats to come.74%want advanced data collection capabilities,such as web scraping or AI-based collection tools like chatbots.Financial authorities seek to push the envelope,with an expressed desire for prescriptive analytics,task auto-mation,advance

152、d image processing,predictive analytics and advanced collection techniques.30|STATE OF SUPTECH REPORT 2022Figure 9.Underpinning technologies currently desired by FINANCIAL AUTHORITIES to en-able AND ENHANCE supervisory processes(N=134)CAMBRIDGE SUPTECH LAB|31The supervisory stack layer where demand

153、for new suptech is highest is the data processing layer,followed by advanced analytics/collection and advanced business intelligence.Figure 10.Underpinning technologies desired by agencies to enable supervisory process-es,organised by layer of the supervisory stack of the suptech taxonomy(N=134)32|S

154、TATE OF SUPTECH REPORT 2022Given the lack of existing adoption of data processing tools and data products noted in the prior section,it is unsurprising to see a high demand to fill these gaps with advanced image processing(79%),advanced text processing(71%),task automation(81%),and advanced business

155、 intelligence tools(70%).Notably,authorities also want to build on existing technologies,introducing more advanced versions of their analytics and collection tools.3.1.3.Suptech generations 2.0A long-term plan for building suptech capabilities inevitably requires that the organisation sets out the d

156、etailed steps by which its current capabilities will be Figure 11.SupTech generations 2.0(Cambridge SupTech Lab 2022 version)expanded;these must be incremental,realistic transitions which leave the organisation with greater supervisory capability than before,even in the relatively short term.BIS(201

157、9)provided a four generations framework to demonstrate how each technology-enabled element of the supervisory process might evolve through time.This framework now serves as an ontology to frame the primary research here on enabling suptech,wherein we extended the suptech generations framework to inc

158、lude a generation zero of manual processes and an additional layer of data products,as presented in Figure 11 and referenced throughout the survey results in this section.CAMBRIDGE SUPTECH LAB|33Suptech efforts remain in the experimentation stage.A vast major-ity of technologies used in the super-vi

159、sory stack of surveyed authorities fall into the first(1G)or second(2G)generation of suptech,which mainly support descriptive and diagnostic analytics.Demand for new suptech is highest for the most advanced 3G and 4G tech-nologies,and decreases with each lower generation.This clear trend validates t

160、he adapted suptech generations framework not only as a descriptor for the current state of suptech but equally as a roadmap for future suptech adoption.Supervisors with existing technology express a clear desire to upgrade,and those without advanced suptech may have opportunities to leapfrog by skip

161、ping earlier generations.34|STATE OF SUPTECH REPORT 2022Figure 12:Underpinning technologies used versus desired by agencies to enable supervi-sory processes,organised by suptech generation(N=134)CAMBRIDGE SUPTECH LAB|353.1.3.1.Data collection Regulatory reporting can be challenging and resource-inte

162、nsive for supervised institutions and financial authorities.As regulatory reporting has become increasingly complex,authorities face challenges in collecting delayed and poor-quality reporting data,which can,in turn,impact their supervisory ability(FCA 2020).63%of authorities collect data through 1G

163、 web-based portals or bulk uploads.Financial authorities collect this data periodically in the form of standard reporting templates.Hence,their focus has been on creating templates rather than using the data to construct analytical reports.A large proportion of the agencies indicated they use a comb

164、ination of manual,bulk(web)uploads and automated reporting to collect data.57%of the surveyed financial authorities have 2G application programming interfaces(APIs).In a move to facilitate more efficient data flow between supervised institutions and supervisors and thus lower the costs,more than hal

165、f indicated having automated their data collection process and have developed an API that allows institutions to submit data.While web-based portals and APIs can support the submission of large amounts of structured data,they are ineffective for unstructured data,such as social media and annual repo

166、rts data.Only 38%of the respondents have 4G advanced collection techniques,which suggests an opportunity for the development of AI-based chatbots,web scraping and data streaming technologies.Most supervisors collect data in formats that necessitate manual processing.Supervisors who are collecting da

167、ta via unstructured digital formats(21.9%),structured flat files(18.9%)and tabular data templates(27.5%)are going to be required to do at least some manual processing,whereas less than one-third of respondents(30.1%)use formats that are machine-readable by default.Figure 13.Data formats in which sup

168、ervisory data is collected by financial authorities,ordered from least(top)to most(bottom)machine-readable(N=74)36|STATE OF SUPTECH REPORT 20223.1.3.2.Data processingEfficient and reliable mechanisms for ensuring quality in data management are fundamental to the supervisory process.The data manageme

169、nt cycle has two main tasks related to data processing(BIS 2019):1.Validation,which refers to the quality control checks of completeness,correctness and consistency of formatting and calculation as per reporting rules2.Consolidation,which involves the integration of data from multiple sources and in

170、 varying formats Many financial authorities still validate data manually through time-consuming and error-prone 0G manual spot checks or 1G spreadsheet-based formulas.The effectiveness of financial supervision relies on efficient data management to provide timely,adequate and accurate data,covering

171、various facets of financialinstitutions business as well as integrating,where necessary,relevant macro data about the markets and the economy.A large proportion of the survey respondents still validate data manually through time-consuming and error-prone manual spot checks(28%)or spreadsheet-based f

172、ormulas(20%),which are unsuited for working with large data sets.The importance of upgrading these methodologies and tools to deliver higher-quality data cannot be overemphasised.A good model cannot overcome inaccurate data,and good-quality data is better than more data.It is,therefore,critical that

173、 financial authorities transition to automated data validation as this ultimately reduces human errors,builds up databases needed for performing analytical work,and enables scarce human resources to be reallocated to more judgement-based work.3.1.3.3.Data storageIncreased regulatory requirements and

174、 the growth of the digital economy have led to a huge increase in available data The integration of 4G suptech to aid in the processing of this big data has been noted as a challenge for supervisors both to prepare for and integrate beyond the pilot phase for some time(BIS 2015b).This trend continue

175、s today,and is likely driven by the challenges(with mitigation strategies)detailed in Section 4 of this Report.An intermediate 3G solution to address the rising cost of big data storage is to use cloud technology,enabling greater and more flexible storage,mobility capacity and computing power.Financ

176、ial authorities have subsequently increased the type of data collected for supervisory purposes and reap the benefits of lower costs and increased storage capacity.However,challenges remain for supervisors in the adoption of cloud computing.First,financial authorities perceive fundamental limitation

177、s and risks when weighed against the current policy,organizational,technical,and legal structures(BIS 2018).Secondly,there are perceived operational risks,cyber risks,dependency and vendor lock-in,data sovereignty,concentration(TC 2020).Finally,even if those perceived risks and limitations are suffi

178、ciently mitigated,there is a need for stronger oversight of regulatory data stored in the cloud(FSB 2020).Financial authorities must take these risks in context and compare not against perfection or an idealized state,but rather against the status quo(HBS 2020).Notably,when treated in this manner,th

179、e risks of not moving to the cloud could be higher than making the move(TC 2020).Many agencies use 1G file-based storage mechanisms or 2G on-premise relational databases.CAMBRIDGE SUPTECH LAB|37Despite the benefits of cloud computing,only a minority of respondents have adopted scalable storage solut

180、ions such as cloud computing(41%)or advanced document management systems such as data lakes and other warehouse techniques(41%).The majority of financial authorities still store their data centrally,or have fragmented,disjointed data management through spreadsheets,desktop databases or paper records

181、.3.1.3.4.Data analyticsSuptech applications support financial authorities in analysing data that are increasing in volume and variety,streamlining processes to drive efficiency and generating intelligence to identifying risks,trends and outliers that might have been missed previously.26%of responden

182、ts conduct 1G manual analysis of supervisory data.These financial authorities rely predominantly on relatively rigid and simplified spreadsheet models for data analysis.Analytical activity is dominated by 2G descriptive and diagnostic analysis.63%of financial authorities leverage these applications

183、that are used to search and summarise historical data to identify patterns or meaning,including automated statistical summaries and the data feeding into dashboards and data visualisation tools.3G predictive analytics is present for 38%of financial authorities,but 4G prescriptive analytics are only

184、adopted by 5%.These tools enable advanced analysis of historical data to create statistical models to predict future events,values,facts or characteristics and then to prescribe an optimal response.This process may include recommendation engines(tools where the prediction is an optimal value or acti

185、on)and employ machine learning(computerised,iterative optimisation of the aforementioned statistical models).3.1.3.5.Data products26%of respondents employ 1G static charts and metrics,while 28%have basic 2G dynamic dashboards.This investment in basic reporting tools to report statistical snapshots a

186、nd key performance indicators(KPIs)is consistent with the substantial investment in diagnostic and descriptive tools at the analytics layer.Only 11%have adopted 4G advanced business intelligence tools.To extract the most meaningful and actionable insights from data,authorities have started to invest

187、 in AI-enabled dashboards that leverage big data tools to allow for numerous analytical operations.38|STATE OF SUPTECH REPORT 20223.1.4.Supervisory areasSuptech initiatives cluster mainly around the areas of consumer protection and prudential supervision.Consumer protection supervision is a relative

188、ly new area of focus for financial authorities.While the growth of fintech and its offerings bring new opportunities,especially for the underserved and marginalised market,it may unintentionally place a greater financial burden and risks on these vulnerable customers.Financial authorities are mindfu

189、l of these inherent risks.They are making deliberate efforts to build trust and confidence in these products,such as expanding their mandates to include responsibilities that some once considered conflicting with the stability mandate,for example,consumer protection,competition and financial inclusi

190、on.Recent advancements in data and technology,such as NLP and real-time monitoring,also support market conduct supervision as they present new opportunities for supervisors by enabling greater qualitative analyses.To the contrary,prudential supervision needed to ensure the safety and soundness of th

191、e banking system is a traditional,core component of many financial authorities mandates.As highlighted above,numerous efforts are being dedicated to upgrade the tools for data collection and validation that are foundational for prudential supervision.The analysis of the regional thematic focus revea

192、led that suptech initiatives that support financial inclusion monitoring are particularly prevalent in sub-Saharan Africa(59.4%)compared with other regions.Figure 14.Thematic areas of financial authorities suptech initiatives(N=134)CAMBRIDGE SUPTECH LAB|393.1.5.Enabling factorsFigure 15.ENABLERS OF

193、data CAPABILITES FOR SUPERVISORY AGENCIES(N=119)There is an unmet desire for data teams,data sharing and data synthesis as a foundational part of the data capabilities of supervisory agencies.Only 25%of respondents who felt a data analysis team was important had a team that was currently active/oper

194、ational.An additional 65%expressed a desire,plan,or ongoing development of a dedicated team.Of the respondents who identified external data sharing as necessary for enhancing regulation and supervision capabilities,only 21%had this core component of DRSI in place and operational.Notably,50%of respon

195、dents who consider data synthesis to be necessary said the technology was desired but not planned,while only 10%had this technology currently operational in their organisation.This high ratio of expressed desire to operation solutions was similar across EMDEs and AEs.3.1.6.FundingFunding for financi

196、al authorities suptech initiatives comes mainly from the financial authorities themselves.Suptech solutions for insurance supervision(80%)and prudential supervision(74.1%)are the least dependent on external funding sources.Only financial inclusion(40.7%)and competition supervision(36.4%)are funded p

197、rimarily by external sources.40|STATE OF SUPTECH REPORT 2022Figure 16.Suptech funding sources,with supervisory areas presented in order of most to least prevalent(N=48)CAMBRIDGE SUPTECH LAB|413.1.7.GovernanceWith the advent of suptech,financial authorities are placing significant focus on evolving f

198、rom ad hoc initiatives driven from within the supervision department and implemented by the IT team to more strategic investment in roles such as Chief Data Officer(CDO),Chief Technology Officer(CTO),and centralised data science units to drive or support suptech implementation.Predictably,the survey

199、 responses show that suptech initiatives are driven by supervision departments(58%)or IT Figure 17.Who is leading THE suptech initiatives(N=40)Who is leading or will lead this SupTech initiative?*Supervision DepartmentOtherChief Technology Officer-IT DepartmentChief Data Officer-Data Analytics Depar

200、tmentResearch/Statistics DepartmentGovernors/Executive Office-Chief of StaffOperations Department58%43%38%35%28%15%10%departments(38%).In a few instances,multiple departments or functions shared the responsibilities for suptech initiatives,a common combination being the supervision department,IT dep

201、artment and data analytics office leading different aspects of the suptech initiative.For example,while a given suptech initiative may be owned and driven by the business unit(supervision department),the data analytics department may be responsible for data strategy,quality and governance,and the IT

202、 department for the technology infrastructure.42|STATE OF SUPTECH REPORT 2022Over one third of the surveyed financial services authorities have designated Chief Data Officer leading suptech efforts.Of note,more than one third of the survey respondents(35%)have a dedicated centralised office reportin

203、g to a CDO who is either solely responsible for the suptech efforts or works with other functions to support the suptech initiatives.This signals the financial authorities growing interest in adopting data-driven approaches to support the supervisory process.Suptech leadership differs across income

204、level and type of financial authoritiesRespondents from financial authorities in AEs reported that suptech efforts are primarily led by a CDO or a data analytics department.In EMDEs,this is still mainly led by the supervision department.Similarly,the leadership of suptech within capital markets,secu

205、rities and investment instruments primarily lies with a CTO,while for most central banks,this charge is led by the supervision department itself.Data is a strategic priority for several financial authorities.Creating a formal data strategy can accelerate data capacities,increase institution-wide buy

206、-in and coordinate action and the support of an authoritys senior management.For example,the FCA developed its first data strategy in 2013 and later updated it in 2020 and 2022 as its data journey transformed over the years,from focusing on how they collect and manage data to supporting its continue

207、d transformation journey towards becoming a digital and intelligence-led institution.Data strategies are typically followed by creating a CDO who leads the development and management of internal data policies and governs data access and management.CAMBRIDGE SUPTECH LAB|43Figure 18.Data strategy matu

208、rity,by financial authority type(N=45)Based on the survey responses,data is a strategic priority for an increasing number of authorities,with most respondents either having a data strategy in place(56%)or in development(33%).While data is a critical factor for central banks,only 43%have a data strat

209、egy in place,and a further 39%are currently developing their data strategy.Financial authorities are actively interested in creating dedicated units to lead strategic data initiatives.The development of formal procedures to govern data access is especially critical for financial authorities as it mi

210、nimises data mismanagement,cybersecurity and information security risks and ensures that the insights derived from data analytics(suptech)solutions support data-driven policymaking.As such,about 46%of the respondents have set up a central unit with a chief data officer who is responsible for their d

211、ata strategy.This group primarily consisted of authorities from EMDEs(61%).The remaining respondents leverage existing units,such as the IT department(44%),to lead their data analytics initiatives.44|STATE OF SUPTECH REPORT 2022Figure 19.Who is leading data strategy initiatives(N=40)CAMBRIDGE SUPTEC

212、H LAB|453.1.8.GenderThere is a great opportunity for suptech solutions to support financial authorities in collecting and analysing granular/disaggregated gender data.Without the collection of national sex-disaggregated,supply-side operational data offering a clear picture of the situation and compa

213、rison between men and women on their access,usage,and quality indicators of inclusive finance,it will be extremely challenging to close the gender gap in financial inclusion and achieve the related economic benefits(AFI 2020).Globally,financial institutions are in different stages of collecting sex-

214、disaggregated data.Very few institutions currently use this data to identify and highlight barriers to womens financial inclusion,whether it be policy-related or awareness and understanding of available financial service products.Survey results are consistent with research findings as most responden

215、ts(70%)stated that they do not have a strategy around financial gender-disaggregated data.In breaking down the overall 21%that are currently operating gender data strategy by type,central banks report slightly greater progress(24%)than other types of authorities(17%).Figure 20.Gender strategy maturi

216、ty,for example,as a standalone workstream,as part of overall data strategy,as part of financial inclusion strategy(N=43)46|STATE OF SUPTECH REPORT 2022Figure 21.Outcomes that financial authorities suptech initiatives have supported(N=88)CAMBRIDGE SUPTECH LAB|473.1.9.OutcomesSuptech can catalyse a ri

217、sk-based supervisory approach that can adapt quickly to a constantly evolving environment.The post-financial crisis reforms,the impact of Covid-19 and the advent of new technologies require that supervisors establish proportionate,risk-based approaches underpinned by efficient data management.Risk i

218、ndicator dashboards,centralised data warehouses for supervisory reports and early warning systems are some tools that are now entrenched in several supervisory agencies worldwide(BIS 2018).Aligned with the above,when asked what they considered the primary outcomes of embracing successful suptech ini

219、tiatives,most respondents pointed to enabling/improved risk-based supervision leading to better identification and measurement of risk(82%),improved and consistent data collection(74%),and increased efficiencies in the use of resources by the reallocation of staff away from manual tasks(68%).At the

220、same time,data and reports submitted by supervised institutions are among the sources of information used most widely by supervisors to inform supervisory activities.Based on the survey results,supervisors also acknowledged suptechs potential to enhance regulatory reporting,with 61%of respondents st

221、ating that a key external outcome of suptech was more efficient information flow between providers and supervisors.Capital market,securities and investment instruments supervisors have seen more substantial internal outcomes than central banks and other supervisory agency types.Respondents in this c

222、ategory have seen more impact in using internal resources more efficiently(82%,versus 63%for central banks and 58%for others)and in greater internal supervisor coordination and information flow(74.1%,versus 61.2%for central banks and 66.7%for others).Financial authorities of all types in advanced ec

223、onomies have seen more substantial internal outcomes than those in emerging markets and developing economies.These authorities have noted greater improvement in scope,accuracy,consistency and timeliness of collected information(85%in AEs versus 67%in EMDEs)as well as greater internal supervisor coor

224、dination and information flow(79%in AEs versus 58%in EMDEs).The impact of suptech solutions generally meets expectations,with few exceptions.Cybersecurity supervision is the only area where respondents report significant levels of low impact(14.3%).The top categories for high impact are competition

225、monitoring(100%),consumer protection and market conduct supervision(79.5%)and AML/CFT/PF(69%).48|STATE OF SUPTECH REPORT 2022Figure 22.Actual vs expected impact of suptech solutions,with supervisory areas presented in order of most(top)to least(bottom)prevalent(N=88)CAMBRIDGE SUPTECH LAB|493.2.Suppl

226、y:Sourcing solutionsSuptech projects can be developed in-house by a team of the financial authority itself,collaboratively with external vendors,or a combination of the two.This strategic decision is usually based on whether the financial authority has the resources or technical capacity to invest i

227、n developing the solutions in-house.When these prerequisites to build in-house are not met,financial authorities must engage with vendors:doing market research to source,running competitions,engaging in pilots,and ultimately conducting formal procurement.Figure 23.How suptech is built,with superviso

228、ry areas presented in order of most(top)to least(bottom)prevalent(N=91)3.2.1.Sources of suptech appsMost suptech solutions are sources externally,either from vendors that develop custom solutions,or as purchased off-the-shelf software.Areas where off-the-shelf and other purchased software were most

229、prevalent are payments(28.6%),cybersecurity(22.2%)and competition monitoring(21.4%).The only supervisory areas where suptech is primarily built internally50|STATE OF SUPTECH REPORT 2022or with internal consultants are climate/ESG risks(72.0%),AML/CFT/PF(60.2%),and capital markets,securities,and inve

230、stment instruments(57.0%).3.2.2.The vendors business caseBased on what we have observed through desk research and direct engagement with many of the 73 vendors listed in the Cambridge SupTech Labs online Vendor Database,for most of them suptech is not the main business,but rather an activity pursued

231、 as complementary to the provision of regulatory technology(regtech)solutions to the more scalable and profitable financial industry,or as part of their broader provision of technologies Figure 24.How strongly suptech factors into surveyed vendors business models(N=6)Figure 25.How DO YOU expect this

232、 suptech component of your business model will have shifted two years from now?(N=6)to a number of industries.So far,we have identified only a few who have made suptech central in their offering and business model.Suptech is currently a strong business case for the suptech vendors that we interviewe

233、d,who expect it to further grow over the next two years.When asked how strongly suptech factors in their business model,all the vendors responded that suptech was either their core business or a key business offering.This suggests that the vendors see suptechs value proposition.They further reported

234、 that they foresee a growing demand for suptech applications in the coming years.Very strongly-Suptech is at the very core of our offeringStrongly-Suptech is an important component among several lines of businessSomewhat-We pursue suptech opportunities as part of a broader strategyMinimally-Suptech

235、act as a loss leader for new markets when necessary,but is not an active pursuitCAMBRIDGE SUPTECH LAB|51Figure 26.Layers of the supervisory stack within which vendors engage with financial authorities(N=6)Figure 27.Suptech funding sources for vendors(N=6)3.2.3.Offerings by focus areaSuptech solution

236、s provided by surveyed vendors focus primarily on the data collection,data processing and data analytics layers of the supervisory tech stack.3.2.4.FundingFrom a vendor perspective,suptech applications are mainly funded by the financial authorities themselves but are also often supplemented by grant

237、 funding.52|STATE OF SUPTECH REPORT 2022CHALLENGES TO UPTAKE4.CAMBRIDGE SUPTECH LAB|534.1.1.ImplementationLimitations in budget,data quality and technical skills remain significant barriers to implementing suptech.Despite the efforts of supervisory authorities to enhance supervisory processes throug

238、h technology,various challenges have been encountered in developing and using suptech applications.As they embark on their modernisation journey,authorities are becoming cognizant of the challenges associated with the digitization of their processes and methodologies.Research has outlined some of th

239、ese challenges,including limitations in data quality,lack of transparency in data,lack of management support and buy-in,increase in cyber security risks in an automated suptech environment,lack of adequate expertise,algorithmic biases,third-party dependencies and legacy systems(BIS 2018,BIS 2019).Th

240、e Report sheds light on the perception and experience of the agencies dealing with these issues.Supervisory authorities reported that budgetary constraints(58%),data quality issues(57%),limited staff with data analytics capability(54%),legacy IT systems(49%),and limited staff IT skills top the list

241、of internal challenges they encounter when developing,deploying,and maintaining suptech solutions.They clearly pointed at challenges related to data analytics and tech development,and the need for capacity building in those two areas.In addition,when it comes to the extenral factors,one third of the

242、 respondents-led by agencies in EDMEs-highlighted challenges when coordinating data sharing with4.1.Challenges:financial authorities54|STATE OF SUPTECH REPORT 2022Figure 28.Challenges faced by financial authorities in developing suptech,grouped by internal and external factors(N=95)external stakehol

243、ders(29%average,38%in EDMEs).Only 7%of these authorities with suptech initiatives lack management buy-in,and only 14%lack a suptech strategy or roadmap,further highlighting these as prerequisites to engaging with suptech.The analysis of responses across different income levels shows that financial a

244、uthorities in AEs and EDMEs report facing similar challenges in the digital transformation of their supervisories processes and capabilities.Different types of financial authorities face different kinds of challenges.While the same challenges appear across all types of financial authorities,their pr

245、evalence differs.Across agencies,budget is the main challenge,but capital markets,securities,and investment instruments supervisory authorities are more impacted.75%of capital markets,securities,and investment instruments supervisory authorities report internal budgetary constraints(versus 49%of cen

246、tral banks and 58%of others).CAMBRIDGE SUPTECH LAB|55Figure 29.Challenges faced by financial authorities in developing suptech,grouped by INCOME LEVEL(N=95)56|STATE OF SUPTECH REPORT 2022For central banks,the challenges are primarily related to internal resistance to change.29%of central banks repor

247、ted internal resistance to breaking data siloes(versus 21%for capital markets,securities and investment instruments supervision and 8%for others).Earlier generations of technology tend to capture and store data in siloes,making it difficult for authorities to gather insights from the data.While big

248、data architecture,such as AI,has the potential to address these challenges,they require data expertise and a data-driven culture(BIS 2019).A lack of this expertise can lead to resistance and pushback from the staff.24%of central banks reported internal cultural resistance to change(versus 21%for cap

249、ital markets,securities,and investment instruments supervisory agencies,and 8%for others).Often,one of the main reasons for suptech implementation is a lack of stakeholder engagement and poorly planned change management.This ultimately leads to resistance by staff.As the financial authorities adopt

250、even more sophisticated technologies or suptech applications,they might find themselves lacking the capacity or skills required,for example,for data analytics or AI/ML applications or lack of understanding of the new processes.The absence of this understanding could lead to a lack of trust in the ap

251、plication results,leading to pushback by staff.This challenge calls for a culture change within the agencies to enable teams to work together and adopt an agile approach.This is especially critical as they begin to engage with the new entrants into the financial sectors who have a different culture

252、from the agencies and operate in different ways.For capital markets,securities,and investment instruments supervisors,challenges tend to be related to upgrading their existing systems and processes.39%have challenges with external coordination with other organisations to access their datasets(versus

253、 25%of central banks and others).36%report underdeveloped/inadequate internal IT infrastructure,like inability to use the cloud(versus 27%of central banks and 25%of others).For other supervisors,the uniquely prominent challenges are with IT systems.75%report challenges related to internal legacy IT

254、systems(versus 40%of central banks and 57%of capital markets,securities,and investment instruments supervisors).75%report insufficient staff with IT skills(versus 38%of central banks and 50%of capital markets,securities,and investment instruments supervisors).Overcoming the common challenges of all

255、financial authorities and the particular challenges for financial authorities of each income level and agency type will help authorities realise the benefits of wider suptech adoption.4.1.2.Data lifecycleThe top challenges for financial authorities are data collection,followed by a lack of analytica

256、l skills and automated data processing.The top five challenges are delays in data submission(16.8%),human error in data collection(15.3%),inadequate analytical skills(13.4%)and incomplete(11.9%)or messy(10.9%)data received due to lack of automated processing,4.1.3.ResourcesFinancial authorities seek

257、 support to build staff skills required to develop and implement suptech solutions.In recognition of the challenges they face CAMBRIDGE SUPTECH LAB|57in accessing talent with adequate skillsets and ensuring that their staff have the right skills to use any suptech applications,financial authorities

258、are investing in building their internal capacity.36%of the surveyed respondents have already undertaken capacity-building activities to support the implementation and use of suptech.A further 45%are interested in supporting their staff to build their skills and knowledge on suptech.Supervisors in e

259、merging markets and developing economies express more interest in capacity-building programmes than those in advanced economies.Figure 30.Data lifecycle challenges along the supervisory stack(N=74)60%of authorities from EMDEs responded that they were interested in undertaking capacity-building progr

260、ammes to support their suptech initiatives,versus only 14%of authorities in AEs.Activities that authorities are undertaking to improve their staff capacity include training programs to enhance technical and digital skillsets and investments in building and fostering a digital culture within the agen

261、cies.For examble the ECB has introduced a comprehensive digital training curriculum to promote a culture of innovation and build knowledge and understanding of suptech.Over 600 supervisors across Europe recently completed an introductory six-week training programme on AI and how it relates to superv

262、isory work.(ECB 2022)58|STATE OF SUPTECH REPORT 2022Figure 31.Engagement in capacity-building programmes in the context of suptech(N=44)Figure 32.Engagement in capacity-building programmes in the context of suptech,segmented by income level(N=44)CAMBRIDGE SUPTECH LAB|59This is part of a broader hub-

263、and-spoke innovation model to foster agile collaboration and the joint development of suptech solutions where innovation teams from the ECB and national supervisors pool their knowledge and contribute to the overall goal of digital transformation.These teams are composed of experts from various func

264、tions(for example,IT,supervision and statistics)with diverse skill sets.Assessing existing capabilities to undertake and successfully implement suptech solutions is an important step.This not only enables financial authorities to map existing resources and gaps but also helps them understand and est

265、imate future capability requirements.Top areas of support sought by financial authorities are training,funding,and technical assistance.When asked what areas of support they required,most respondents pointed to the same top five,including technology training(83%),data science training(81%),funding f

266、or design and development of suptech tools(69%),cybersecurity training(67%)and technical assistance for data analytics(64%).A key consideration for financial authorities as they adopt and adapt suptech is their capability level regarding skills,talents and resources.Nearly all central banks primaril

267、y seek training on technologies,whileFigure 33.Areas of support sought by financial authorities(N=42)60|STATE OF SUPTECH REPORT 2022securities supervisors seek a mixture of training on cybersecurity,a digital platform for conversations,and technical assistance and funding for suptech development.Res

268、ponses were segmented by agency type to understand the top areas of support for each type.Central banks tend to seek training more actively,most prominently on technologies(92%).Capital market,securities and investment instruments supervisors seek training on cybersecurity(83%),technical assistance

269、to build applications(67%),a digital platform for peer conversations(75%),and funding for the design and development of suptech(92%).Financial authorities in emerging markets and developing economies express more need for support than those in advanced economies.In particular,tech training(93%),data

270、 science(90%),suptech process(70%)and technical assistance on data analysis(80%)and conducting a diagnostic(60%)were in demand for EMDEs.The demand for digital tools was similarly strong across both categories.In terms of funding,financial authorities in EMDEs,in particular,focused on funding for th

271、e design and development of suptech solutions(77%),while those in AEs more often focused on funding for hiring(42%).4.1.4.InfrastructureFinancial authorities face considerable challenges in digital infrastructure in their jurisdictions.The top two challenges are limited knowledge/expertise(cited by

272、63%of respondents)and funding/resource constraints(57%).Legacy IT systems(reported by 49%),a lack of capabilities(48%),poor quality or insufficient data(44%),and the availability of technology(42%)are also common challenges.As financial authorities develop strategies on how to upskill,train and buil

273、d capacity internally for data collection,analysis and management,they can also consider the following strategies:Develop comprehensive(digital)curricula to help staff build both technical and soft skills required to thrive in an innovative environmentRecruit and retain digitally skilled staff such

274、as data scientistsCollaborate with different stakeholders to facilitate knowledge transfer and peer-to-peer learningTap into external technology solutions vendors with vast knowledge of IT and suptech systems,who can assist financial authorities in dealing with rapid changes in technology and overco

275、ming limited in-house technical skills and resources.4.2.Challenges:vendorsWhile there is an increasing demand for suptech applications from financial authorities,vendors face obstacles due to the procurement process(83%),dealing with siloed teams and multiple stakeholders within the financial autho

276、rities(50%),lack of visibility into the needs of financial authorities(50%),and lack of adequate funding sources(50%).Additional challenges in providing suptech solutions relate to engaging with financial authorities(in particular technical capacity of supervisors and long sales cycles),developing s

277、uptech applications(compliance with data protection mechanisms,digital infrastructure limitations,insufficient access to historical data,compliance with cybersecurity requirements,lack of global data standards and limitations to transfer technologiesCAMBRIDGE SUPTECH LAB|61from early adopters to the

278、 rest of the authorities)and the expanding the vendors suptech portfolio(mapping their technological offerings to supervisory use cases,the lack of global scalability of their solutions,legal and regulatory restrictions,and the relatively small size of the suptech market).Procurement processFor fina

279、ncial authorities to achieve their goals of transforming supervisory processes and tools,they must access the right capabilities,capacities,services and products.These are sometimes provided internally by civil servants,but in many cases need to be obtained from the private sector through public pro

280、curement processes,which are flagged by most suptech vendors(83%)as the main challenge they face to engage with the agencies.Public procurement is a sensitive domain that must be carried out efficiently and to high standards in order to safeguard Figure 34.TOP FOUR challenges faced by vendors in exp

281、anding a suptech portfolio(N=6)the public interest.Designed to promote accountability,integrity and effectiveness in the management of public budgets,procurement processes often“enact arduous procedures regulated by long and complex legal frameworks,which may limit the capacity for innovative ideas

282、to be implemented,or even considered.While the strict procedures surrounding public sector procurement aim to protect public money,they often generate perverse incentives,delay processes and could ultimately compromise the quality of service delivery.Such complex public procurement systems and proce

283、sses represent a major hurdle to SME participation in public procurement markets,as such companies are disproportionately affected by these factors,due to limited financial,technical and administrative capacities.”(OECD 2019)Numerous obstacles deter smaller,innovative outlets to participate in publi

284、c bids,including“a lengthy and overly complex contracting process,a lack of clarity on how to connect with agencies,and a sense that newcomers have little chance to win contracts over incumbents(BCG and Eastern Foundry 2017).62|STATE OF SUPTECH REPORT 2022It is important to note that most(88%)suptec

285、h vendors listed in the SupTech Marketplace are MSMEs.Micro(1-10 employees)and small(11-50 employees)enterprises make up 45%of the marketplace,while medium enterprises(51-500 employees)constitute another 43%.Only 12%of vendors are large firms(501+employees).Moreover,during the past several years the

286、re has been notable consolidation occurring within the nascent suptech market.Of the 74 suptech vendors listed in 2018 by the RegTech for Regulators Accelerator(RA),20 of them have now been acquired by larger firms.Rather than maturing independently and providing competitive solutions,these vendors-

287、who represented 27%of the recorded market in 2018-were since absorbed into larger competitors.Given the expressed procurement challenges and the asymmetric effect these challenges have on MSMEs,one plausible inference is that the challenges faced by small vendors,primarely due to procurement process

288、es,may be impacting competition and innovation in the suptech space,and conversely resolving procurement issues could increase the volume and variety of competitive offerings available to financial authorities seeking suptech solutions.Financial authorities seem to underestimate the negative impact

289、of public procurement procedures on vendors(Table 29).The concern expressed by interviewed vendors resonates with our experience,and is supported by an extensive body of literature that identifies the rigidities in government procurement rules among the main causes of failure of public sector digita

290、l technology projects(Dunleavy and Carrera 2013,World Bank 2016).Siloed teams/multiple stakeholders within the financial authoritiyAnother relevant challenge for vendors is dealing with siloed teams and multiple stakeholders within the financial authorities.When it is not clear“who the client is”and

291、 vendors are not aware of all the departments and units that are involved in the development and implementation of a modernization project,they end up losing time and burning budget while dealing with redundancies and delays in engaging with the appropriate people.Lack of visibility into the needs o

292、f financial authorities More and better visibility into the needs of financial authorities would allow vendors to tweak their products to serve more specific needs,and to develop a business pipeline that would make their businesses more viable and attract investments.Insufficient funding Vendors may

293、 face challenges in covering the costs of business development in a market that,at first,may offer little scale,and where payments can be deferred and delayed as part of the procurement cycle.The Cambridge SupTech Lab partners with financial authorities and technology vendors to co-create and deploy

294、 cutting-edge,scalable suptech applications.The Lab furthers the capacity of financial authorities to drive their engagement with technology providers by helping them develop proofs of concept and technical specifications(see the Digital SupTech Diagnostic Tool),identify off-the=shelves solutions an

295、d developers outlets that are serving this market(see the SupTech Vendor Database)and also acting as a broker that curates and facilitate the collaboration between agencies and vendors.The Labs Application Foundry-which builds largely on the successful experience of the RegTech for Regulators Accele

296、rator(RA)-accelerates the development of ground-breaking suptech applications by detailing the technical specifications,de-risking procurement for all parties,providing project management support and hands-on technical assistance,introducing an agile approach to the collaboration between financial a

297、uthorities and technologists.CAMBRIDGE SUPTECH LAB|63CASE STUDIES5.The following suptech case studies serve to complement the insights derived from the survey and presented in the analyses above,and to ground the findings in a practical context.This select set of case studies has been systematically

298、 drawn from across the supervisory stack,representing each layer of the“tech”side of the suptech taxonomy presented in section 1.2.2.While each case study involves multiple layers(e.g.,a data collection solution often involves data processing),they are classified by the layer at which the primary in

299、novation occurred.A final case study includes a solution that focuses on a holistic approach,innovating at all layers of the stack in parallel.A database of solutions is available in the Cambridge SupTech Labs Suptech Marketplace.64|STATE OF SUPTECH REPORT 20225.1.Data collection:Bank of England tra

300、nsforming data collection from the UK financial sectorIn 2019,the Bank of England(BOE)conducted a review of the future of the UKs financial system,and what it might mean for the BOEs agenda,toolkit and capabilities over the coming decade.The Future of Finance report(BOE 2019)identified some critical

301、 issues in relation to the efficiency and effectiveness of data collection,pointing at policies and processes that made the collection of data from regulated firms costly,time consuming,relatively inflexible,and often redundant for both the BOE and the industry.The report suggested that a number of

302、underlying factors may contribute to these issues:Heterogeneity in firms dataFor any given product or transaction,different firms might hold and describe equivalent data differently.This makes it hard for the BOE to write a set of reporting instructions that are unambiguous to all firms.In turn,this

303、 can lead to pain points for firms in interpreting instructions and locating data,which has the potential to cause long timelines and quality issues for the BOE.Heterogeneity of the Banks data needsReports are designed to address specific use cases.For instance,the BOE often requires data to be aggr

304、egated in ways that make reports hard to repurpose.This leads to more requests for new reports or breakdowns of existing reports than would otherwise be the case.It also leads to redundancy in the reporting process,as firms need to re-assemble the same underlying building blocks in different ways fo

305、r different reports.Duplication of processes across firmsMany elements of the production of reports are common across firms.This raises the possibility that further centralising some processes could reduce duplication and improve the systems efficiency.The review sought ways to decrease the burden o

306、n industry and to increase the timeliness and effectiveness of data in supporting supervisory judgements.Because the BOE determines what information is required of regulated firms,it has a significant influence on their data governance and management.Therefore,the report pointed at the potential for

307、 the BOE to better support the firms own use of data,making them more productive and competitive.The BOE committed to conduct-in consultation with the industry-a review to explore a transformation of the collection,hosting and use of regulatory data over the next decade,identifying ways to decrease

308、the burden on the firms and increase the timeliness and effectiveness of data in support of supervision.Five key challenges relating to the efficiency and effectiveness of how data is collected,which can be found in different jurisdictions across the world,were identified(BOE 2021):Complexity,legacy

309、,and strategic planning The two main sources of complexity are:i)on the firm side,the data for reports can come from various types of legal entities,business lines,and operational systems,and ii)on the government side,different authorities ask for similar data with slightly different definitions,acr

310、oss multiple reports,at different breakdowns.In addition,the legacy of decisions made in the past have created a complex reporting landscape that has not been adapted to the current data needs.This is reflected in a legacy of manual,siloed processes,and outdated,fragmented operational systems.If the

311、 authorities are overly focused on meeting short-term objectives and lack strategic planning,they will not tackle these legacy issues.CAMBRIDGE SUPTECH LAB|65Value and collection rationaleData collection contributes to improving decision-making and making it evidence-informed.However,an agency can f

312、ind it hard to estimate the value of reporting.Interpretation Industry participants expressed that understanding reporting instructions was one of the greatest sources of(avoidable)cost of the data collection process.The difficulties they mentioned include finding the latest version of the instructi

313、ons,locating all the relevant documents,navigating the BOE website hosting the instructions,and understanding the instructions as they are written in over-complex legal language.1 1 2 2 3 3 Common data inputs Modernising reporting instructionsChanges to the reporting architectureDeveloping common da

314、ta inputs at a more granular level would provide a defined way for firms to record certain data(for example,data elements for individual mortgages)or capture the key elements in a common input layer.This could provide a more consistent cross-firm foundation from which to build reports,reducing costs

315、 and improving speed and quality.Common data inputs could also form the basis of a move to modernise how the Bank writes reporting instructions.This could include moving from our current natural language approach towards more precise instructions for se-lecting and transforming the data of interest.

316、Doing so could reduce the cost and time it takes for firms to respond to new requests.Common data inputs could also support different architecture solutions,such as a pull data collection model.A pull model would allow the Bank to query certain data held within firms and generate reports on demand.T

317、his could improve the speed and flexibility of reporting while reducing the marginal cost to firms of responding to new questions.Finding and sourcing data Due to their legacy systems and the complexity of data requested,at times industry participants found it to be challenging to locate or source t

318、he required data.Reconciliation and data quality Complexity and legacy issues made resolving data quality problems unduly difficult.The BOE formulated its vision for data collection,which is that“The Bank gets the data it needs to fulfil its mission,at the lowest possible cost to industry”,and devel

319、oped its transformation plan for data collection addressing three areas:66|STATE OF SUPTECH REPORT 2022The BOE-FCA Joint Transformation ProgrammeTo help deliver this reform,in 2021 the BOE and FCA set up a Joint Transformation Programme,in collaboration with the industry.During each phase,with an it

320、erative and pragmatic approach,the Programme aims to deliver a series of use cases,defined as a collection,set of related collections,or an aspect of a data collection.Within each phase,each use case,in turn,passes through a discovery and design stage,and then an implementation stage(Beta),where sol

321、utions are developed and tested for delivery.Phase 1:Discovery and designThe use cases selected for this phase were:Commercial real estate CRE data(BOE use case),with a focus on improving the quality and coverage of commercial real estate data available to the Prudential Regulatory Authority(PRA)and

322、 the BOEs directorate for Financial Stability Strategy and Risk.Quarterly derivatives statistical return Form DQ(BOE use case),aiming improve data on the derivative asset and liability positions of the UK financial sector.This data ultimately feeds into the UKs balance sheet compiled by the Office f

323、or National Statistics.Financial resilience survey(FCA use case),looking to formalise a post-pandemic ad-hoc collection of select data points from FCA firms,used for prudential risk monitoring.Through a process of workshops and discussions between subject matter experts from the industry and regulat

324、ors,the delivery teams identified similar issues across the different use cases with seemingly similar root causes.Some of these issues were:User experience challenges(such as users finding it difficult to find the right information they need to prepare reports,including the context and rational for

325、 the collection)Difficulties firms face in understanding and interpreting reporting requirements Issues firms and regulators face in providing and receiving feedback on data quality A lack of understanding by firms on the impact the data has on the regulators/supervisors(such as how the data is used

326、 to improve decision-making)Concerns that the same or similar data was being collected across multiple collections.The identification of similar issues across different use cases supported the Programme hypothesis that data collection processes can be redesigned to achieve better effectiveness and e

327、fficiency.Phase 2:Implementation(ongoing)The use cases for this phase are:Commercial Real Estate(CRE)data,building on the findings of phase one,which confirmed that the current CRE data the agency receives is inadequate,fragmented,and burdensome to collect.Phase two will focus on exploring business

328、practices and processes,mapping user journeys,creating a project roadmap and creating problem statements.Strategic Review of Prudential Data Collection(SRPDC),to reduce the cost of data production and reporting for firrms and could deliver more value to both industry and agencies.Retail Banking Busi

329、ness Model Data that the FCA currently collects across a range of retail banking products and segments,and is critical to support the FCAs competition objective and is reused by other FCA business units.The data is currently collected ad-hoc,without an integrated design that meets the needs of all o

330、f the data users.CAMBRIDGE SUPTECH LAB|67 Incident,Outsourcing and Third-Party Reporting(IOREP),which the BOE,the Prudential Regulatory Authority(PRA)and the FCA use to ensure the operational resilience of the financial sector,relying on data with low quality and consistency.In addition,these activi

331、ties have been undertaken:Data Standards Review(ongoing)As part of their vision for increased development and adoption of common data standards throughout the financial sector,BOE and FCA set up the Data Standards Committee,which commissioned a review of data standards.Banking Data Review(PRA,ongoin

332、g)The PRA has announced the launch of a Banking Data Review through a discussion paper(BOE 2022b)that outlines the increase in the scope of PRAs policy-making responsibilities and the expected impact of the Financial Services and Markets Bill.This is initiative is being run separately by the PRA,how

333、ever,it is complementary to the work being carried out within the Joint Transformation Programme.Transition to Bank of England Electronic Data Submission(BEEDS)portal In 2020,the BOE Data and Statistics Division(DSD)announced the plan to move the collection of statistical data to the Bank of England Electronic Data Submission(BEEDS)portal(BOE 2020).BEEDS is an online application that enables firms