(AGENPARL) – ven 10 marzo 2023

Testo Allegato: Conference “The New Frontiers of Digital Finance”
Paolo Angelini
Rome, 10 March 2022
1 â??Traditionalâ?
I would argue that this is probably the sector least a�ected, thus far, by the latest
technological developments that inspire today’s seminar. The adoption of digital
technologies for �nancial market infrastructures started with the dematerialization of
transactions at a�ordable costs, thanks to their ability to e�ect a huge increase in
e�ciency (a rise in the velocity of circulation of central bank money). The next step was
to eliminate counterparty risk in securities transactions via the adoption of delivery
allowed investors to use real-time data for algorithmic and high-frequency trading,
which in Europe now account for about 70 per cent of all equity trades. Nowadays,
re�ecting a widespread dematerialization process, almost all �nancial assets are in
digital form and traded through electronic platforms; all wholesale payments are
Algorithmic trading implements order and trade decisions electronically and autonomously (without
orders are submitted and trades executed at high speed, usually microseconds, and a very tight
This said, the innovation pipeline is by
no means empty. In the time allotted I shall review some recent developments
of the “traditional” type, and devote some remarks to the debate on DLT-based
In Europe, the RTGS world has been evolving rapidly. In 2018 the Eurosystem
and their clients – corporates and households alike – to transfer funds within seconds,
around the clock, 365 days a year. The Eurosystem charges banks well below one cent per
transaction. The project’s quality is widely acknowledged, as witnessed by the Sveriges
Riksbank’s decision to join TIPS with its currency, and by the intention to follow suit
enhance
and modernize the services o�ered by TARGET2. The new system features the ISO
20022 messaging standard (as in T2S and TIPS), a centralised liquidity management
that will make it more cost-e�ective and e�cient,
and can facilitate payments in
several currencies, if other central banks decide to join the system.
A third example is
the creation of the Eurosystem Collateral Management System (ECMS), a multi-year
project that started in December 2017 and is expected to go live in April next year.
The ECMS will replace the current fragmented and decentralised structure for the
management of collateral – de facto a patchwork of domestic systems. It will allow
participants to assess and move collateral on a domestic and cross-border basis, and
seamlessly across the euro area.
added to TARGET2, implementing a safer and more e�cient securities settlement.
It should also be acknowledged that, at least in some speci�c instances, the digital revolution has been
a mixed blessing. Consider e.g. algorithmic and high-frequency trading: in normal conditions they may
risk aversion and market instability, or generate â??�ash crashesâ? â?? episodes of sudden and large price
A centralised liquidity management tool will function via the so-called “Main Cash Account” (MCA)
that participants can open with a national central bank. This account will be linked to the participant’s
MCA will also o�er a dashboard for a centralised overview of liquidity positions and advanced liquidity
management tools, with a higher level of automation. The liquidity held on dedicated cash accounts
will be considered for minimum reserve purposes without the need for the accounts holder to transfer
2 The
the novelties brought about by Distributed Ledger Technologies (DLTs) like blockchains.
international fora that are working to de�ne global regulatory standards for these markets,
including the Financial Stability Board, the Basel Committee on Banking Supervision and
proposing principles and benchmarks for supervised intermediaries and entities falling
Coming to DLTs use for �nancial market infrastructure, two main types of claims
have been made. According to early-hour proponents of DLTs, the technology has the
potential to do away with �nancial intermediaries and their trust-building role: the very
key features of the technology – the distributed nature of the ledger, the “impossibility”
for any single user to tamper with the records once validated – pave the way to a world
with a greatly diminished role for intermediaries. This â??decentralized �nanceâ? (DeFi) view
emphasizes the self-tending nature of DLTs, i.e. their ability to work without a subject in
charge of managing the system. In this context, reference is often made to the so-called
A di�erent view holds that DLTs can improve the e�ciency of �nancial market
infrastructure, reduce the time needed for reconciliation and back-o�ce activities, o�er
high performance and programmability thanks to the use of so-called ‘smart contracts’.
In this case the emphasis is not on the DLT governance but rather on the bene�ts that
can stem from the technology, independently of whether or not a �nancial intermediary
is in charge of it.
Let me express some scepticism about the former view. One way or another,
counterparty risk would �nd its way into a fully decentralized large scale trading and
This is because these systems, to be scaled up, need in practice
some centralized infrastructure (e.g. exchanges) to provide various services (e.g.
storage of tokens and cryptographic keys, trade, conversion into other tokens or hard
…). In other words, DeFi, beyond a certain operational scale, is forced to
reintroduce the infrastructure that it was supposed to eliminate. And several episodes
have shown that this infrastructure is prone to various risks (cyber and fraud risk in
Communication by Banca dâ??Italia on Decentralized Technology in Finance and Crypto-assets
, Rome,
See e.g. the insightful discussion by F. Schar, â??Decentralized Finance: On Blockchain- and Smart
Federal Reserve Bank of St. Louis REVIEW, second quarter 2021,
particular), just like the traditional one. Once this is acknowledged, DeFi schemes lose
The second view cannot be easily dismissed. It is undisputable that DLTs have useful
Bank of Italy (through our Milano Hub, the innovation centre created to support the
digital evolution of Italian �nancial system) launched a speci�c call for proposals on DLT
applications in �nance, insurance and payments. We received
57 projects, submitted by
82 participants. We are currently evaluating them. While the majority of projects comes
from Italian entities, a non-negligible share comes from other European countries and
So the issue becomes an empirical one: is large scale DLT adoption in
I believe these are important questions. While it is not for central banks to answer
them, I think that they provide a strong motivation for the central banking community’s
and settlement systems. For instance, TARGET2 o�ers e�cient and secure settlement in
central bank money, but it would be unable, without the development of ad hoc interfaces
in DVP mode using tokenized commercial bank money. These exciting developments
o�er an opportunity to test the DLT and get practical feedback about its bene�ts (and
disadvantages). In my view, however, DVP in central bank money is no longer optional
The list of snags could go on. Permissionless DLT systems must continuously generate enough fees
for anonymous validators to ensure the integrity of the ledger in the absence of a central authority.
Ultimately, this creates signi�cant ine�ciencies and risks for users. For example, when the value of
Luna went to zero in the recent Terra-Luna collapse, the incentive for miners to validate transactions
evaporated (as they were remunerated in Luna units), and the blockchain stopped for several
hours. A similar problem could materialize for any blockchain managing securities. See P. Lee, â??Has
tokenizationâ??s time �nally come?â?, Euromoney, March 03, 2023. In DLTs based on proof of work,
mining is highly energy- and hardware-intensive; this has tended to concentrate mining activity, with
the attendant risk of collusion and forks. DLTs relying on proof of stake tend to concentrate decisions
on few holders of governance tokens, reintroducing the governance problems that are typical of
Banca d’Italia is actively promoting research on the topic of ‘smart contracts’ including their implications
on cyber-security. A
Memorandum of Understanding
was recently signed with Roma Tre University
This is just one of many possible examples of the ferment around new technologies in �nance.
According to a market sounding recently conducted by the ECB among banks, �nancial market
infrastructure operators and new �ntech �rms, about 70% of participants expect a signi�cant industry
uptake of new technologies, such as DLTs, in the next 5 to10 years; views are mixed on the relative
merits of DLTs compared with existing technologies, as well as on the expected timing of DLT adoption.
Potential use of new technologies for the settlement of wholesale �nancial transactions in
bring us back to the seventies, when the term Herstatt risk (a form of counterparty risk)
For this reason, the central banking community is working to develop a form of
that such solution is a necessary condition â?? although certainly not a su�cient one â?? for
The Eurosystem is currently exploring two main architectural models: a wholesale
CBDC service fully based on DLTs, with tokenized central bank money issuance (so
“cash on DLT”) and an interface component integrating DLT platforms with current
With the caveat that this is genuine work in progress, the �rst model (cash on DLT),
however ingenious, presents operational complexities and could increase the overall
Furthermore, the presence of two segregated wholesale
one for each tokenized market/security) would stoke risks of liquidity fragmentation,
jeopardizing the optimization e�ort undertaken by the Eurosystem with the T2-T2S
system in central bank money and one or more external private DLT platforms handling
tokenized digital assets. Compared to the �rst model, such a solution would not create
reduce adaptation costs and implementation time, and would likely be less vulnerable to
Banca d’Italia has successfully experimented a version of this second model,
centred on TIPS. In a nutshell, the prototype provides a DLT-agnostic protocol to
synchronize the asset-leg and the cash-leg of a tokenized asset, making an instantaneous
payment transaction possible on a 24/7 basis. The paper documenting
3 Concluding
To conclude, I believe that, concerning large-scale DLT adoption for �nancial
requires actions by central banks, and is still under investigation. Solid evidence
about actual bene�ts of DLT adoption for �nancial market infrastructure (in terms of
M. Nardelli and C. Oliviero, â??
Systems series, no. 26, July 2022.
increased e�ciency, security,
â?¦) is not yet available. Actually, a prominent project to
move a securities exchange to the DLT technology was recently brought to a halt by the
Australian Securities Exchange (ASX), after seven years of work and a sunk cost of about
â?¬250 million. On the other hand, according to market estimates, global spending on
blockchain-based solutions, while currently negligible, has been steadily growing (from
$ 1 billion in 2017 to 19 billion in
2022). Market interest in asset tokenization is also
DLT developments may have been held back by the lack of a sound legislative
framework. In the EU, an important stimulus could come from the DLT Pilot Regime
Regulation, which aims to foster innovation in the processing of tokenized securities.
Banca d’Italia is cooperating with the Italian Ministry of Finance and Consob for a swift