Monetary Policy and the Economy Q2/21

Call for applications: Klaus Liebscher ­
Economic Research Scholarship

Please e-mail applications to scholarship@oenb.at by the end of October 2021.
Applicants will be notified of the jury’s decision by end-November 2021.

The Oesterreichische Nationalbank (OeNB) invites applications for the “Klaus ­Liebscher Economic Research Scholarship.” This scholarship program gives outstanding researchers the opportunity to contribute their expertise to the research activities of the OeNB’s Economic Analysis and Research Department. This contribution will take the form of remunerated consultancy services.

The scholarship program targets Austrian and international experts with a proven research record in economics and finance, and postdoctoral research experience. Applicants need to be in active employment and should be interested in broadening their research experience and expanding their personal research ­networks. Given the OeNB’s strategic research focus on Central, Eastern and Southeastern Europe, the analysis of economic developments in this region will be a key field of research in this context.

The OeNB offers a stimulating and professional research environment in close proximity to the policymaking process. The selected scholarship recipients will be expected to collaborate with the OeNB’s research staff on a prespecified topic and are invited to participate actively in the department’s internal seminars and other research activities. Their research output may be published in one of the department’s publication outlets or as an OeNB Working Paper. As a rule, the consultancy services under the scholarship will be provided over a period of two to three months. As far as possible, an adequate accommodation for the stay in Vienna will be provided. 1

Applicants must provide the following documents and information:

  • a letter of motivation, including an indication of the time period envisaged for the consultancy
  • a detailed consultancy proposal
  • a description of current research topics and activities
  • an academic curriculum vitae
  • an up-to-date list of publications (or an extract therefrom)
  • the names of two references that the OeNB may contact to obtain further information about the applicant
  • evidence of basic income during the term of the scholarship (employment contract with the applicant’s home institution)
  • written confirmation by the home institution that the provision of consultancy services by the applicant is not in violation of the applicant’s employment contract with the home institution

1 We assume that the coronavirus crisis will abate in the course of 2021. We are also exploring alternative formats to continue research cooperation under the scholarship program for as long as we cannot resume visits due to the ­pandemic situation.

Nontechnical summaries

in English and German

Nontechnical summaries in English

Distributed ledger technologies for securities settlement – the case for running T2S on DLT

Alfred Taudes, Jakob Hackel, Wolfgang Haunold, Hannes Hermanky

This paper proposes to use distributed ledger technology (DLT) to make the Eurosystem’s TARGET2-Securities (T2S) system for securities settlement more flexible. Specifically, we present a framework where building blocks of ­information on transactions are replicated and validated by several computers. Apart from technological issues we also address regulatory compliance, performance, cost efficiency and risk. The system we propose is a federated system comprising central banks and central securities depositories, enabling the participating European central banks and central securities depositories to conduct securities settlement according to regulatory requirements. The role of the central banks is to maintain the cash accounts; provide the infrastructure which allows for securities issuance, ­settlement processes, auto-collateralization, corporate actions such as dividend payments and more. The central securities ­depositories provide and maintain securities accounts, offer notary services for issuers, perform corporate actions, and carry out settlement. Ledger updates are based on a fully automated process, allowing central securities depositories to conduct settlement while central banks retain regulatory control. The system processes various types of settlement instructions for different security classes using specialized smart contracts, meaning programming code that executes agreements automatically on DLT infrastructure. This provides the flexibility to settle various types of digitally ­represented assets and to define novel workflows. In particular, the study shows that this also allows for variable ­settlement times. The proposed DLT architecture also enables central banks and authorized actors to conduct status checks at a granular level and in real time. Furthermore, comparisons show that the proposed system, when ­implemented with current distributed ledger technology, meets the daily performance goals of T2S. The federated structure of the system would also increase the resilience of operations. Overall, the study proposes one way to approach implementing securities settlement via a novel technology, while conforming to regulatory and operational requirements. The paper aims to contribute to ongoing discussions on the suitability of DLT technology for traditional settlement processes.

The share of zombie firms among Austrian nonfinancial companies

Christian Beer, Norbert Ernst, Walter Waschiczek

Aggregate productivity and economic growth may be reduced by “zombie firms” – weakly performing companies that, instead of exiting the market or being restructured, manage to continue operating over an extended period. We present first results on the incidence of such zombie firms for Austria, based on three definitions that capture different aspects of this phenomenon. Our results suggest that the share of zombies fell substantially between 2009 and 2018, across all industries and firm sizes, albeit to different degrees. The drop of the zombie share was particularly strong for highly indebted enterprises. Still, at the end of our observation period zombie firms continued to have less favorable risk ­characteristics than non-zombie firms, in particular a distinctly higher probability to default. Further findings were obtained with simulations keeping short-term interest rates stable over the period under review. Under this ­assumption, the zombie share would have remained roughly constant as well. The difference between the observed and the ­simulated zombie share is particularly pronounced for real estate-related industries, more highly indebted firms, and larger ­companies. Finally, the data show that zombie status is not irreversible. Among those firms for which financial ­statements information is available for the entire observation period, most firms manage to exit from zombie status at some point. One big unknown that remains is, of course, how these patterns may have changed as a result of the current pandemic, because our data do not go beyond 2018.

Austria’s labor market during the COVID-19 crisis

Christian Ragacs, Lukas Reiss

The impact of the COVID-19 crisis on Austria’s labor market has been huge and a lot heavier than during the Great Recession of 2009 in terms of the increase in unemployment and the drop in employment. Key metrics show that the decrease in employment was broadly in line with the euro area average and that the increase in unemployment went hand with an increase in long-term unemployment and the average duration of unemployment. The generous ­short-time work scheme rolled out by the government prevented a turn for the worse and also lessened the downward pressure on average wages induced by the strong decrease in average hours worked per employee in 2020. While manufacturing or construction were hit as well, the tourism industry was affected most by the crisis, contributing to a relatively stronger increase in unemployment in provinces with a higher tourism-related share of employment. Younger employees and especially foreigners were also relatively more affected by the increase in unemployment, while employees with tertiary education were relatively less affected. Labor supply, while losing momentum, did continue to grow in 2020, while it had stagnated during the Great Recession.

Nontechnical summaries in German

Modell zur Nutzung innovativer Technologien (Distributed Ledger) für die Abwicklung von Wertpapiergeschäften in Europa

Alfred Taudes, Jakob Hackel, Wolfgang Haunold, Hannes Hermanky

In dieser Studie werden Überlegungen angestellt, die aktuelle Eurosystem-Plattform zur Abwicklung von ­Wertpapiergeschäften – TARGET2-Securities (T2S) – durch Umstellung auf DLT-Technologie flexibler zu machen. DLT steht für Distributed Ledger Technology. Vereinfacht gesagt ist damit gemeint, dass die einzelnen Bausteine aller Transaktionen auf vielen Computern verteilt gespeichert werden. Konkret behandelt die Studie die Frage, ob das ­Wertpapierabwicklungssystem auch auf DLT-Basis die gesetzlichen Vorschriften erfüllen kann, und wie es um ­Leistungsfähigkeit, Kosteneffizienz und Risiken steht. Das vorgeschlagene Modell zur Interaktion der Zentralbanken und Wertpapiersammelbanken ist ein föderales System, das den teilnehmenden europäischen Zentralbanken und ­Wertpapiersammelbanken die Wertpapierabwicklung weiterhin im Einklang mit den gesetzlichen Vorschriften ­ermöglicht. Laut Modell führen Zentralbanken die Geldkonten und stellen die Systeme bereit, die für die Ausgabe von Wertpapieren, die Wertpapierabwicklung, die automatische Besicherung der Wertpapiergeschäfte und die ­Durchführung anfallender Dividendenzahlungen und dergleichen benötigt werden. Die Wertpapiersammelbanken wiederum verwahren die Wertpapiere in Wertpapierdepots, übernehmen im Zuge der Wertpapierausgabe Notariatsfunktionen und wickeln die Wertpapierverrechnung sowie Dividendenzahlungen etc. ab. Der Prozess, auf dem die DLT-Datenaktualisierung im Zuge der Wertpapierabwicklung durch die Wertpapiersammelbanken basiert, ist ­vollautomatisiert, und die zentralbankseitige Regulierung des Systems bleibt gewahrt. Das System ist auf die Ver­arbeitung verschiedener Abwicklungsaufträge für unterschiedliche Wertpapierklassen auf Basis sogenannter intelligenter Verträge ausgelegt; das heißt, die Abwicklung läuft DLT-basiert wie programmiert ab. Damit können digital ­darstellbare Werte auf Basis neuer Workflows sicher den Besitzer wechseln. Das neue System eignet sich weiters dafür, den ­Abwicklungszeitpunkt variabel zu bestimmen. Die vorgeschlagene DLT-Architektur ermöglicht darüber hinaus den Zentralbanken und autorisierten Teilnehmern Statusabfragen auf sehr detaillierter Ebene und in Echtzeit. Vergleiche zeigen, dass das DLT-System auch die für T2S geltenden Ziele in puncto Leistung erfüllt. Die föderale Struktur wirkt sich zudem positiv auf die Robustheit des Systems aus. Insgesamt zeigt die Studie eine Möglichkeit zur Nutzung einer neuen Technologie für die Wertpapierverrechnung unter Einhaltung der gesetzlichen und operativen Vorgaben auf. Zweck der Studie ist es, einen Beitrag zur aktuellen Diskussion über den möglichen Nutzen der DLT-Technologie für herkömmliche Abwicklungsprozesse zu leisten.

Entwicklung des Anteils von „Zombiefirmen“ am österreichischen Unternehmenssektor

Christian Beer, Norbert Ernst, Walter Waschiczek

Produktivität und Wirtschaftswachstum können unter sogenannten Zombiefirmen leiden – also unter Unternehmen, deren Betrieb eigentlich eingestellt oder umstrukturiert werden sollte – die aber weiter aktiv bleiben. Im Rahmen der vorliegenden Studie wurde erstmals der Anteil der Zombiefirmen am österreichischen Unternehmenssektor analysiert, und zwar auf Basis von drei Definitionen, um den Begriff Zombiefirmen möglichst gut eingrenzen zu können. Die ­Analyse zeigt insgesamt einen deutlichen Rückgang von Zombiefirmen zwischen 2009 und 2018; je nach Branche und Firmengröße war die Entwicklung aber unterschiedlich. Besonders deutlich war der Rückgang unter den stark ­verschuldeten Unternehmen. Risikotechnisch betrachtet schneiden die Zombiefirmen am Ende des Beobachtungszeitraums letztlich aber noch immer schlechter ab als gesunde Firmen, insbesondere im Hinblick auf die ­Wahrscheinlichkeit einer etwaigen Insolvenz. Ferner wurde simuliert, wie sich der Zombieanteil entwickelt hätte, wenn das Zinsniveau im Analysezeitraum auf dem Stand von 2009 geblieben wäre. Unter dieser Annahme wäre auch der Anteil der Zombiefirmen weitgehend stabil geblieben. Am größten ist der Unterschied zwischen dem beobachteten und dem simulierten Zombieanteil bei den Unternehmen in der Immobilienbranche, bei den vergleichsweise stärker verschuldeten sowie bei den größeren Unternehmen. Schließlich lässt sich von Entwicklung der Bilanzdaten (soweit durchgehend vorliegend) ableiten, dass der Zombiestatus meistens nicht unumkehrbar ist. Da die Daten der vorliegenden Studie nur bis 2018 reichen, bleibt die Frage offen, wie sich diese Ergebnisse im Lichte der aktuellen Pandemie verändert hätten.

Österreichs Arbeitsmarkt während der COVID-19-Krise

Christian Ragacs, Lukas Reiss

Die Auswirkungen der COVID-19-Krise auf den österreichischen Arbeitsmarkt waren enorm. Sowohl der Anstieg der Arbeitslosigkeit als auch der Rückgang der Beschäftigung fiel deutlich stärker aus als während der Rezession 2009, ­wobei der Beschäftigungsrückgang weitgehend dem Euroraum-Durchschnitt entsprach. Mit dem Anstieg der Arbeitslosigkeit erhöhte sich auch die Langzeitarbeitslosigkeit und die durchschnittlichen Dauer der Arbeitslosigkeit. Die von der Regierung eingeführte großzügige Kurzarbeitsregelung verhinderte eine noch schlechtere Entwicklung und ­verringerte auch den Abwärtsdruck auf die Durchschnittslöhne, der vom starken Rückgang der durchschnittlichen Arbeitszeit pro Arbeitnehmer im Jahr 2020 ausging. Branchenweise wurde der Tourismus verglichen mit dem ­verarbeitenden Gewerbe oder dem Baugewerbe ungleich stärker von der Krise erfasst, was zu einem überdurchschnittlichen Anstieg der Arbeitslosigkeit in Bundesländern mit einem höheren Beschäftigungsanteil der Touristikbranche beitrug. Vom Anstieg der Arbeitslosigkeit waren jüngere Arbeitnehmende und insbesondere ausländische Personen vergleichsweise stärker betroffen, Arbeitnehmende mit Hochschulbildung hingegen relativ weniger. Das Arbeitskräfteangebot erhöhte sich zwar weniger stark als in den Vorjahren, die Tendenz blieb aber im Jahr 2020 auch in der Krise steigend, im Gegensatz zur stagnierenden Entwicklung während der Rezession 2009.


Distributed ledger technologies for securities settlement – the case for running T2S on DLT

Alfred Taudes, Jakob Hackel, Wolfgang Haunold, Hannes Hermanky 2 , 3
Refereed by: Fabian Schär, University of Basel

With a view to developing the Eurosystem’s TARGET2-Securities (T2S) system further, we propose a system based on distributed ledger technology (DLT) that covers all major T2S settlement functionalities and investigate it with regard to regulatory compliance, performance, cost efficiency and risk. The system we propose is a federated system comprising European central banks and central securities depositories (CSDs) as node operators. The role of the central banks is to maintain the cash accounts; provide regulatory-approved “smart contract factories” defining workflows for securities issuance, lifecycle management and matching, settlement, auto-collateralization and corporate actions; and perform the oversight function. The CSDs maintain securities accounts, offer notary services for issuers, perform corporate actions, and carry out settlement. CSD nodes collect settlement requests from external trading and clearing systems, forward them to other CSDs for cross-border settlement, bundle them into transaction blocks and prepare the blocks for settlement. The ensuing ledger updates occur via a fully automated consensus process between the central banks. In T2S on DLT, specialized smart contracts provide the flexibility to settle a range of digitally represented assets, define novel workflows – and allow for variable settlement times. Rather than having to conform to a uniform settlement time of T+2, participants can choose among smart contracts that settle within seconds or longer periods of time. This feature is expected to reduce capital costs and, given the DLT-based enforcement of settlement discipline, settlement failures. Apart from conforming to the current regulatory requirements, the DLT framework also enables the central banks and authorized actors to conduct status checks at a granular level and in real time. Furthermore, comparisons with similar use cases and benchmarks show that the use of current DLT solutions would allow to meet the current daily performance goals of T2S. Preliminary cost ­estimates based on available public information indicate that the proposed system could be built and operated efficiently. The federated structure would also support the resilience of operations given the high number of backup nodes.

JEL classification: E44, G21, G23, K22

Keywords: distributed ledger technology, securities settlement, smart contracts, TARGET2-­Securities

1 Introduction

In securities transactions, ownership transfer is typically a three-stage process: first, a deal is established via brokers using a trading engine. Second, transactions are cleared through a central counterparty (CCP) to limit counterparty risk and provide netting benefits. Third, the deal is settled by central securities depositories (CSDs) through the coordinated updating of cash and securities accounts (see ­figure 1).

Figure 1 is a visual that illus-trates post-trade processes in the securities leg of cur-rent transactions as a three-stage process, as described in the text. Source: Pinna and Rutten-berg, 2016.

1.1 Overview of TARGET2-Securities

In Europe, securities settlement in central bank money was harmonized and centralized with TARGET2-Securities (T2S), the common platform launched by the Eurosystem in June 2015. Before T2S, cross-border securities settlement in Europe was costly and cumbersome due to different settlement practices among countries and complex cross-border settlement procedures. As outlined by the ECB on its website, “T2S lays the foundations for a single market for securities settlement and thus contributes to achieving greater integration of Europe’s financial market. It does this by:

  • making it easier for investors to buy securities in other EU Member States
  • reducing the cost of cross-border securities settlement
  • increasing competition among providers of post-trade services (i.e. clearing and settlement services) in Europe
  • pooling collateral and liquidity, meaning that banks no longer need to keep these in various locations and can quickly move them to where they are needed
  • reducing settlement risk and increasing financial stability by using central bank money for transactions on the platform.”

T2S matches settlement instructions submitted by central securities depositories and other directly connected T2S actors and coordinates the updates of the securities and cash legs of the transactions. T2S securities accounts are managed by the 21 participating CSDs while cash accounts are managed by the participating central banks. Of the six transaction categories handled by T2S, delivery-versus-payment (DvP, simultaneous transfer of central bank money and securities) accounted for 66.43% of the volume processed in 2019, followed by free-of-payment (FOP, delivery of securities without a simultaneous transfer of funds) with 30.60% of the volume (ECB, 2021a) (see figure 2). PFOD (payment free of delivery) transactions, SRSE (settlement restriction on securities) transactions, LQT (liquidity transfer) transactions and DWP (delivery-with-payment) transactions together ­accounted for 2.97% of the daily average volume. In addition to transactions to exchange securities, T2S handles corporate actions such as payments of dividends and administrative functions like the issuing of new securities.

T2S settles trades in four phases within two days. From 6:45 p.m. to 8:00 p.m., securities are validated, and liquidity is transferred from TARGET2, the real-time gross settlement system operated by the Eurosystem, to dedicated cash accounts in T2S. Then the nighttime settlement cycle is performed from 8:00 p.m. to 3:00 p.m. After a two-hour maintenance window, daytime settlement starts, at the end of which liquidity is transferred back to TARGET2 until 6:00 p.m. The time remaining to the next cycle is used for reporting.

Figure 2 is a visual that illus-trates the components and workflow of the Eurosys-tem’s TARGET2-Securities (T2S) platform. Source: OeNB, 2017

If cash or securities necessary for settling a transaction are not available, settlement fails. To enhance settlement efficiency, T2S provides credit through auto-collateralization, either through “auto-collateralization on flow” (with transactions being secured by the securities that are being purchased) or “auto-collateralization on stock” (with transactions being secured by securities already held by the buyer). If settlement fails despite these measures, T2S retries settlement during a 60-day recycling period. In 2019, the percentage of transactions settled by T2S on the same day was 93.27% (ECB, 2021a). Since 2020, settlement failures have been discouraged by financial penalties and mandatory buy-ins under CSD Regulation (EU) No 909/2014 (EquityClear, 2020).

T2S is owned by the Eurosystem and run by the central banks of France, Germany, Italy and Spain. Following initial development costs of about EUR 584 million, the annual running costs amount to about EUR 102 million. In 2019, T2S processed on average 606,938 securities transactions per day, with a value of EUR 1,106.13 billion. On average, intra-CSD transactions, i.e. transactions within individual countries, represented 99.03% of all trades (ECB, 2021a). As T2S is a central platform, special efforts have been made to ensure robust operation, including rotation between primary and secondary sites. In 2019, the availability of T2S was equal to or above its target of 99.7%.

Notwithstanding major achievements of TS2, including higher security and lower cross-border transaction cost, the following shortcomings remain, as identified by BNP Paribas (2019):

  • “The cost of settlement has not reduced. Indeed, it has increased. Even if this is explained by transaction volumes (lower than predicted at the launch of the project in 2010) and by the project’s high amortization costs, it is disappointing
  • Cross-border settlements (from one CSD to another) are easier and cheaper. However, they still account today for less than 1% of T2S transactions and hence are comparatively insignificant
  • We do not yet have a single European capital market – issuers have continued issuing in the markets that they and their investors know best
  • Competition between CSDs remains limited. Its impact on pricing is unclear, as some CSDs have increased their asset servicing fees. Nor has competition led to consolidation of CSDs
  • Settlement and asset servicing remain interdependent.”

1.2 Previous assessments of DLT use for securities settlement

Distributed ledger technologies (DLTs) enable the distributed storage and sharing of data records. At the same time, DLTs ensure data integrity through applied cryptography and distributed consensus-based validation protocols. Furthermore, they enable the execution of complex transactions via smart contracts, i.e. programming code that executes agreements automatically on DLT computer infrastructure. Due to their many functional features, DLTs have a potentially transformative impact on financial services. Implementation of these technologies is currently being considered for many financial services functions, including payments, deposits and loans, market provisioning, investment management, and insurance.

The potential of DLT and blockchains, a specific type of distributed ledger technology, for securities settlement has been studied by several authors from various perspectives.

Pinna and Ruttenberg (2016) analyze the current securities post-trade landscape and point out that the necessity to reconcile silos of information controlled by different intermediaries results in complex and expensive processes. They find that embracing DLT to increase the internal/cluster efficiency of existing players would not lead to substantial gains as long as current business practices remain ­unchanged. In contrast, a radical transformation of trading, where issuing companies, fintechs and governments would set up a peer-to-peer system in the spirit of Bitcoin, could radically reduce post-trade cost and trading time – but this approach is in conflict with the current regulatory environment. Therefore, Pinna and ­Ruttenberg (2016) propose the collective adoption of DLTs by CSDs and central banks as a third alternative, which would yield substantial benefits within the current regulatory environment.

Analyzing the potentials and risks of DLT for payment processing and securities settlement, Deutsche Bundesbank (2017) finds that DLT promises improvements in transparency and immutability, operational efficiency, security and resilience, independence from intermediaries, and automated contract processing. Yet, these improvements come at a cost: privacy may be at risk if current encryption technology becomes unsafe as technology improves, a network with different node types may be more prone to attacks, and links to real assets represented on a DLT have to be established by an intermediary.

The Committee on Capital Markets Regulation (2019) argues that public blockchains like Bitcoin are not suited for real-time settlement of securities transactions. They recommend collaborative efforts of existing stakeholders to improve current systems based on permissioned blockchains, pointing out the integration of the cash leg and reversibility of transactions as important issues. Similarly, Chiu and Koeppl (2019) state that, “For policymakers and regulators, three key themes emerge from our analysis. First, to ensure DvP, it is important to link digital ledgers for asset ownership and payments together to support atomic trades. Second, the feasibility of using a blockchain for settlement depends on a sufficient volume of transactions, high enough costs for tampering with the blockchain (possibly in the form of fines) and a limited default exposure. Here, regulation and supervision could play a role to ensure such conditions. Finally, in the case of a permissionless blockchain, coordination to adjust its design might prove difficult. Here, the regulator can help to coordinate the different participants to reach agreement.”

Mainelli and Milne (2016) report the outcome of a series of interviews and focus group meetings with professionals working in post-trade processing and the provision of mutual distributed ledger services. Respondents argued that DLTs for securities transaction processing would need to be permissioned, and that substantial reengineering of business processes is needed to reap the benefits of a transition to DLT.

Another factor to be taken into account is settlement time, as it determines the collateral and regulatory capital necessary to cope with counterparty risk. This would imply that the settlement time should be as short as possible. Yet, a too short settlement cycle may require dealers to pre-fund their trades or to borrow the securities they need to settle. In 2013, for instance, the Moscow Stock Exchange transitioned from real-time settlement to T+2, citing security borrowing costs as the key rationale for this move (see Pavliva, 2013). It therefore makes sense to implement flexible settlement times, letting market participants choose this parameter. However, Khapko and Zoican (2020) argue that flexible settlement speed coupled with mandatory borrowing can lead to an inefficient race to shorter-than-optimal settlement cycles, excessive security borrowing activity, and economic rents for security lenders. They find that this tension is reduced by flexible failure-to-deliver penalties that depend on the cost of borrowing securities, disciplining security lender competition and allowing for real-time settlement. In a DLT based settlement system, such features can be implemented in the smart contracts that handle settlement in a straightforward way (see Szabo, 1994 and 1997, for smart contracts).

The BIS Committee on Payments and Market Infrastructures (2017) provides an analytical framework for analyzing the use of blockchains in payment, clearing and settlement. The framework consists of a method for describing the architecture of a DLT network and a set of questions regarding the efficiency, safety, and market implications of a proposed system.

Parallel to the ongoing theoretical analysis of DLT, several prototypes of DLT systems for securities settlement have been built and studied, the most notable ones being Jasper, Stella, Ubin, Blockbaster and Helvetia (see Bech et al., 2020).

Project Jasper is a collaboration between Payments Canada, the Bank of Canada, TMX, Accenture and R3 (see Bank of Canada et al., 2018). In the project, DvP of equity tokens representing a claim on equity held in Canada’s depository system against cash tokens representing a claim on the Bank of Canada was tested using atomic settlement on the same ledger. It was found that the new process was more efficient and less risky when compared to the existing settlement system.

Project Stella is a collaboration between the ECB and the Bank of Japan (see ECB and Bank of Japan, 2018). The project tested single-ledger and cross-ledger DvP using security and cash tokens with a focus on settlement failures. In the single-­ledger case, trades were found to fail when trading details had not been agreed between parties or when validation of the transaction failed. In such cases, the tokens remain with their owners, exposing traders to replacement cost risk only. In cross-ledger DvP, trades were found to fail if one leg of the transfer could not be delivered, exposing participants to principal risk, too. Hence, the ECB and the Bank of Japan (2018) conclude that an arbitration function on the ledger is needed to deal with such cases.

Based on these findings, Project Ubin (Monetary Authority of Singapore, et al., 2018) served to build a framework for governing settlement processes, including arbitration processes to deal with settlement fails and a recognized market operator for monitoring and facilitating market functionalities. These new processes were found to compress the settlement cycle and to reduce principal risk.

Project Blockbaster by Deutsche Börse and Deutsche Bundesbank (see Deutsche Bank and Deutsche Börse, 2018) served to investigate the performance of DLT for securities settlement using the Hyperledger Fabric system. In these experiments, DLT was shown to fulfill the performance requirements necessary for building real-life settlement systems. Moreover, Deutsche Börse, Deutsche Bundesbank and Germany’s Finance Agency have recently demonstrated that it is possible to establish a technological bridge between blockchain technology and conventional payment systems to settle securities in central bank money with a transaction coordinator in TARGET2 (see Deutsche Bank, 2021).

In Project Helvetia, the BIS, in cooperation with SIX Group AG and the Swiss National Bank, showed that it is possible to provide central bank money to settle securities transactions in a realistic near-live setting using new technologies (see BIS et al., 2020). This exercise confirmed the feasibility of linking up the existing systems and of issuing digital central bank money.

Shabsigh et al. (2020) summarize the findings from DLT prototype projects in settlement as follows: “In general, the DLT prototypes showed that DLT could be viable for post-trade securities processing and all projects concluded that securities settlement is a highly suitable and feasible environment for DLT-based solutions. The experiments showed that different DvP models can be implemented in DLT-based systems. DLT solutions can vary considerably in features and tools, with which a more efficient processing and account method can be designed and customized for improved efficiency and security in specific markets according to market needs.”

Regarding the architectures tested, Shabsigh et al. (2020) state: “In all prototypes, the central bank was given the role of cash instrument provider and could thereby also be the one ensuring DvP requirements. A project assumption appeared to be that securities clearing and settlement systems operate within a market structure close to the current structure—that is, exchanges, dealers, CCPs, CSDs, custodians, and central banks operate in similar or near-similar roles as they do today and in a multilayered registration structure. None of the projects analyzed flatter market structures and DvP processing at the end-investor level or other radical structural changes in the market and associated risks.” With regard to open issues, Shabsigh et al. (2020) mention the study of the impact of real-time 24/7/365 processing on the design of the system, the analysis of liquidity and credit risk in a realistic setting and the possible changes in market structures.

In addition to the experimental project discussed above, two DLT-based systems for securities settlement are at the pre-production stage, one in Australia and one in Switzerland. The Australian Stock Exchange has developed a DLT-based system for clearing, settlement, and securities registration to replace its current system, called CHESS (Australian Stock Exchange, 2020). The new system, which builds on a prototype developed in 2016, only covers the securities leg. Development costs were reported to amount to USD 50 million.

The Swiss DLT settlement prototype, developed by SDX, covers the full securities value chain including order entry, order crossing in the matching 4 engine, DvP settlement on Corda DLT and distributed holdings of intermediated securities/tokenized cash. The reported costs are USD 100 million so far. According to SDX (2019) “Test-cases will showcase the potential of SDX’s riskless trading model, as well as settlement on DLT. Early-stage functionality will cover digital security token issuance as well as live trading and instant settlement. This will include the cash-leg of the transaction embracing the concept of a payment token as well as access to a distributed portal where it would be possible to monitor transactions across specific DLT member nodes.” When moving to the new system, SDX expects costs to decrease due to reduced collateral requirements, lower operational costs, and lower data management costs enabling potentially lower fees per transaction. Further benefits anticipated are an increased asset universe, new primary and secondary markets, a private marketplace for interbank/inter-client trades, real-time information at the client holding level enabled by the link between asset and owner and a significant simplification of corporate event handling.

1.3 Outline

Based on these findings, we present a DLT-based architecture for T2S and describe how T2S would work in this new environment. In the description, we concentrate on delivery-versus-payment (DvP) transactions and corporate actions. The principles discussed here also apply to the other transaction types. Subsequently, we analyze the feasibility and advantages of our design from a technical and economic perspective and with regard to compliance with the regulatory environment/framework based on the framework presented by the BIS Committee on Payments and Market Infrastructures (2017). In the conclusion section, we summarize our findings and outline some topics for future research.

2 Specification of T2s on DLT

2.1 Architecture – understanding the arrangement

Type of DLT: Following the findings in literature, the system we propose is not publicly accessible but private and permissioned. It employs a heavily centralized consensus protocol, based on a predefined (closed) set of consensus-relevant nodes, and restricts read access. Unlike with public DLT solutions like Bitcoin or Ethereum, this provides advantages from a regulatory perspective, as public DLT solutions are incompatible with existing case law (Pinna and Ruttenberg, 2016). Moreover, permissioned DLTs have a higher capacity and fewer restraints than current permissionless blockchains. Public blockchains currently have trade-offs in lower throughput to preserve as much decentralization as possible (see Schäffer et al., 2019). Since permissonless networks are incompatible with current regulation regarding EU-wide securities settlement, our system makes a tradeoff in decentralization to achieve higher throughput than decentralized public solutions. Such a tradeoff necessarily comes with the caveat of loosened immutability, since permissioned networks are by design controlled by the parties that are authorized to establish consensus on the state of the network. Additionally, the permissioned model does not allow regular users to verify transactions or the current state themselves, which is a core premise of the permissionless model as used in some public blockchains.

Some of these conditions might change in the future. Regulation might adapt and become more welcoming to decentralized consensus on a European settlement layer, accepting some level of power over the infrastructure by unknown participants while preserving regulatory oversight and final say over settlement on a settlement layer connected to a permissionless blockchain implementation. Moreover, public and decentralized DLTs are steadily improving their base layers, as well as introducing various solutions of off-chain scaling possibilities, such as Layer 2 implementations. Such Layer 2 implementations allow for a fully secure, slower settlement layer while more scalable implementations handle most of the network load. Notably the Lightning Network is an ongoing attempt to outsource load from the Bitcoin network onto a Layer 2 solution, which handles transactions through a network of bidirectional payment channels (Poon and Dryja, 2016). Most implementations attempt to solve scalability issues by conducting off-chain transactions and only committing limited data as proof to the settlement layer. Some of these implementations use rollups, mainly ZK rollups (which commit bundled transactions through more complex “zero-knowledge proof” technology) and optimistic rollups (which leverage users actively monitoring and reporting on invalidly committed proofs) (see Whitehat, 2018, for ZK rollups and Optimism for an implementation of optimistic rollups). Additionally, “decentralized finance” (DeFi) tools might bring forward newer technological innovations, which might allow for a preservation of decentralization with high throughput. DeFi acts under different constraints and is described in detail by Schär (2021).

Nodes: The DLT design proposed comprises two types of nodes:

  • Central bank nodes: These nodes are authority nodes and designed to manage the infrastructure and develop it further. They have supervisory duties and are responsible for maintaining cash accounts, regulating access to the cash leg and the trading engine, and updating the ledger of transactions.
  • Central securities depository nodes: These nodes are settlement nodes and ­responsible for maintaining securities accounts, issuing new securities, performing settlement, and handling corporate actions. CSD nodes have selected read access to the ledger.

Users: Users send settlement instructions to be relayed by CSDs and receive reports about their holdings and transactions. They do not have access to the ledger but interact with the system via a dashboard which enables them to access messages, send message requests, and select transaction possibilities and currently available methods for settlement. Users, such as commercial banks, can sign their transactions with a key pair. However, direct access to the ledger, which would allow for independent verification by users, would necessitate either a much higher transparency of other users’ actions or highly complex cryptographic methods, which would introduce drawbacks and complexities of their own. In our system, the group of users is still restricted to holders of central bank accounts.

Accounts: We distinguish between user-controlled accounts and smart contract accounts, much like permissionless blockchains such as Ethereum (see Buterin, 2014). Members of central banks, CSDs and user organizations interact with the system via user-controlled accounts, with the control tools being keys for the authorization of transactions with one or multiple signatures, depending on local governance rules. Smart contract accounts contain computer code that sends transactions which constitute ledger updates if included in a block, based on function invocations (see Szabo, 1994 and 1997). For T2S on DLT, we propose smart contracts to ensure that buying and selling instructions are executed (settlement smart contracts), to represent securities holdings and execute corporate actions (securities smart contracts), and to enable auto-collateralization.

Smart contract factories: Smart contract factories enable central banks to offer a dynamic collection of regulatory-compliant building blocks for settlement processes rather than default smart contracts limited to a single type of settlement and security. Smart contract factories serve to generate a range of smart contracts based on predefined standards, a feature already used on public blockchains such as Ethereum by projects such as Uniswap and Authereum. Thus, CSDs can create specific smart contracts to enable transactions with certain assets (e.g. stocks) to be settled in a certain way (e.g. DvP) with certain workflow specifications (e.g. T+1) subject to the prevailing technical and regulatory requirements. In other words, while being responsible for the execution of transactions, CSDs are constrained by the types of assets and workflows available from the smart contract factories managed by the central banks. Allowing for contracts to be made fully flexible without mandatory smart contract factory control might imply too drastic a departure from the way T2S operates today.

Basic settlement process: Figure 3 provides an overview of the proposed system’s consensus and describes the path for a transaction to be included in an update of the ledger’s state, by inclusion in a block.

Transactions are broadcasted from one CSD to the next, with each CSD storing transactions to be processed in a pool until they are settled with finality through inclusion in a block, i.e. a permanent record of transactions, by him or another CSD. Whenever a CSD is chosen by the proof-of-authority (POA) selection mechanism to be next in line for block proposition, he selects transactions outstanding from the pool and then proposes his block to the central banks by broadcasting it. The central banks check the block formally for requirements and sign off if the block was formed validly, or refuse their signature if it was formed invalidly. If a block is not accepted, the proposal mechanism chooses another CSD and starts the process again. Blocks that are signed off as valid are stored as an update to the ledger by all parties, who also delete the now-included transactions from their pool, and the process starts again.

In this setting, national banks cannot suggest that a particular transaction be settled. Nor can they reject a transaction for arbitrary reasons, which would be obvious to everyone else. All of this happens within seconds. Unlike in permissionless systems, which currently establish consensus mostly based on proof-of-work (POW, see Nakamoto 2008), in a POA consensus mechanism all participants are known and identifiable. This restricts the participation in settlement to CSDs and central banks, as current regulation demands. As a necessary drawback, the system is highly centralized when compared to open and permissionless blockchains. Most notably, the framework enables CSDs to censor transactions by not relaying them further or by not including them into the blocks they form. This issue is somewhat mitigated by users being able to send their transactions to a different CSD as well as central bank oversight. Additionally, as central banks have final authority over the infrastructure they can roll back transactions to restore a previous state, either if reconciliation is impossible otherwise or if enough central banks were to collude (“malicious nodes”). For a more formal description of POA and its benefits and limitations see De Angelis et al. (2018).

Figure 3 is a flowchart that shows the transaction pro-cess: Transactions are broadcasted via central se-curities depositories, who are chosen at random by the proof-of-authority con-sensus algorithm. The de-positories check whether a transaction request is new and correct and then bundle broadcasted transactions into a “block” for proposal to the network of central banks. If accepted, a status update is made and the algorithm chooses the next depository; if it is not ac-cepted, the algorithm will choose the next depository. Source: OeNB/Vienna Uni-versity of Economics and Business.

Cash leg: Each central bank has the right to initiate new cash generation against collateral to provide liquidity to the system. When the DLT system is launched, each participating central bank defines their initial cash balance. This balance is prefunded and serves initial funding requirements. To top up this balance, each central bank can initiate self-funding transactions to be included in the next block; much like Bitcoin blocks include “Coinbase” transactions, which allow for the creation of new Bitcoins. Together, all central banks also operate a multi-signatory account, requiring signatures from all parties, to collect cash that is to be removed from the system. Once sent to said address, the respective funds are locked for anyone and can only be unlocked through a unanimous decision of all central banks. These two mechanisms allow for the creation and destruction of cash in the system. Together, they allow for a flexible cash regime where clients in Target2 can deposit cash with their central bank, which in turn creates that amount of cash via a self-funding transaction on the DLT ledger and then sends this cash to the client’s cash account on the DLT ledger. Until TARGET2 has been consolidated with TARGET2-Securities, which removes the requirement to net out end-of-day balances, this DLT process can also be used to create and remove daily balances. When the multi-signatory address has a large enough balance, central banks can use this already existing but locked cash to fund cash accounts again or use it for auto-collateralization by unanimous decision and signature.

Overview of roles and obligations: Figure 4 summarizes the roles and obligations in the system. Central banks act as validators and are obliged to propagate blocks. They also mint/burn cash and are able to lock funds. The central banks cooperatively create smart contract factories, keep the infrastructure up to date by building and updating the protocols, handle permissions of CSDs and approve blocks. The CSDs can create new contracts using the smart contract factories provided by the central banks, and handle transactions and corporate actions. Moreover, it is their job to validate the existence of securities and to propagate transactions. Users initiate and sign transactions.

Figure 4 is a visual listing the roles and obligations as-signed to the three partici-pants of the system: central banks, central securities depositories and users. The central banks are in control of the cash leg and have widespread privileges and control. The central securi-ties depositories control the securities leg of the transac-tions and propose new blocks. Users can only sign or broadcast transactions. Source: OeNB/Vienna Uni-versity of Economics and Business.

2.2 Security smart contracts

Creation: Similar to regular token contracts on permissionless DLTs, such as ERC20 (Vogelsteller and Buterin, 2015), which contain standardized functions that, for instance, allow anyone in the system to check the balance of any user, a T2S DLT ­security smart contract is created by a CSD and contains standardized functions. It consists of building blocks provided by the smart contract factories, which define the possible data structure and entries available to the CSD. This mechanism ensures that newly created securities comply with the relevant regulatory requirements. Factories are used to provide largely harmonized constructs for corporate actions and settlement processes, too. The ability of smart contracts to interact with each other allows for transparent automated processes. Similar security contracts have been proposed for standardization in permissionless blockchains such as Ethereum via EIP/ERC-1400 (Dossa, 2018).