The ‘fintech era’, otherwise known as the ‘industrialisation’ of the financial sector, is a worldwide transformation of the financial sector where new digital technologies will enable greater automation. Artificial Intelligence (AI), which feeds on and learns from the vast amounts of data generated by the financial sector, will be key in automating routine tasks, improving decisions and mitigating risks. AI’s natural language processing systems can read through and begin to unpick regulations and contracts. Yet despite heavy use of digital technologies and computers, the financial industry still has a great deal of potential to be automated, among others at the very core of the financial information processing network. One of the impediments standing in the way of this is missing or incoherent standards for data and algorithms. Professor Wolfgang Breymann and his team are working on one specific but crucial aspect of preparing banks for the fintech era: financial contracts.
The ACTUS approach
Imagine you want to know about the assets that a bank has; how much it will be obliged to pay in the future and the level of risk that the bank is exposed to. This becomes particularly pressing in a liquidity crisis, such as the one sparked by the 2007-8 financial crash: you need to know if the bank can still pay its dues and whether it can still do so if the financial environment worsens. Yet taking full audits of this information – regulators call them “stress tests” – can take weeks if not months, contributing to uncertainty and protracting the crisis. Part of this is because any bank’s balance sheet consists of millions of contracts, many of them steeped in legal jargon. This complexity and diversity of contracts makes the automation of this crucial process very difficult. This is where ACTUS (Algorithmic Contract Types Unified Standards) comes in.
Professor Breymann argues that, although contracts may appear diverse and complex, they can be standardised in a way that would allow machines to understand them. “However, if one abstracts from the legal terms and focuses on the cash flow obligations, the diversity of financial contracts or agreements shrinks drastically. The vast majority of the relevant financial contracts are built on a manageable number of underlying mechanisms. Financial contracts follow a limited number of patterns”, according to the ACTUS website. (http://actusfrf.org)
The ACTUS project (Algorithmic Contract Types Unified Standards) has specifically been designed to enable the full range of financial analyses for risk management and financial regulation. ACTUS is a data and algorithmic standard for the representation of all financial contracts developed and promoted by the ACTUS Financial Research Foundation and is particularly useful for “smart” contracts (financial contracts that perform all functions automatically). It is a real collaborative effort rooted in practice: a precursor developed by Dr Willi Brammertz still provides the basis of Wolter Kluwers’ risk management platform. Teaming up with Prof Breymann was crucial for turning it into the ACTUS technology suitable for a worldwide standard.
One element of the financial crisis was the prevalence of complex financial derivatives such as CDOs (collateralised debt obligations); they repackaged loans in a way that wasn’t transparent. When the subprime mortgage crisis hit, the liquidity crisis that followed owed to that lack of transparency: no one could be sure which banks were exposed to the ensuing risks. By dividing the contracts into specific subgroups based on the structure of their cash flow obligations, which can be understood and ‘read’ more easily by algorithms, ACTUS hopes to ensure that more automated banks can speak to each other in the same language, and regulators can carry out risk assessments faster, more easily and objectively, thus establishing automated contract processing “from cradle to grave” and automated risk reporting at the level of the individual contracts.
The fulfilment of contracts has come back into the spotlight with the development of blockchain technology. The blockchain consists of a distributed ledger of information – such as transactional information – which is maintained and secured by the users (often in exchange for a token payment) through cryptographic technologies, cutting out the middleman who independently verifies that the contract has been fulfilled. Smart contracts not only define the rules and penalties around agreements in the same way that a traditional contract does, but also automatically enforce them. Defining the rules – and verifying that they have been enforced – requires a consistent platform that translates the complexities of the contracts into something machines can understand.
This is where ACTUS data standards come in, not only allowing risk-assessment and financial analysis but also improving transaction processing and risk reporting. Indeed, the startup Ariadne Business Analytics founded by Dr Brammertz and Professor Breymann together with other colleagues, teamed up with Blockhaus to develop an integrated, professional digital banking infrastructure. This will be the first banking platform that integrates traditional and FinTech banking channels while at the same time automating core banking processes such as transaction processing, bookkeeping as well as risk analysis and reporting. Furthermore, the fact that Ariadne was chosen to participate in the Thomson Reuters Labs™ Incubator shows rapidly increasing interest in the ACTUS technology. In addition, the European Commission signalled interest in automating financial reporting by using Blockchain technology and ACTUS.
ACTUS combines the contractual information and the algorithms that translate those contracts into cash-flows, given additional information in form of so-called risk factors that determine the state of the financial environment as, e.g., interest rates, foreign exchange rates, and so on. Then, ACTUS generates the raw results in the form of cash flow streams in a specific data format that’s portable and easy to analyse. In addition to this, the different types of financial analysis such as liquidity and solvency calculations are carried out on top of these raw results. This encompasses income analysis, sensitivity analysis and different kind of risk measures. Notably, preliminary tests and even a proof-of-concept demonstration have already been performed by implementing Monte Carlo simulations (a statistical strategy that relies on repeated random sampling to obtain results), using real data sourced from the European Central Bank and from one of the large US asset managers.
The future and end-goal
Eventually, the ACTUS team aims to have a framework that’s widely used; interest of a large data provider has already been aroused; once the algorithms that represent contractual rules have been tested, and the platform is accepted, an advantage of the platform is that you can use it to calculate the response to arbitrary risk factors. In other words, you can stress-test the response of the system to hypothetical values of interest rates, stock prices, etc. – hopefully allowing for better decision-making and risk analysis, and averting catastrophes. As to the individual banks, ACTUS can undo errors and reduce complexities at the data warehouse level and Smart Contracts will reduce cost of analytics drastically.
I think it is the other way around: Blockchain technology will profit from ACTUS. Or more precisely: ACTUS and BC (or a more general distributed ledger) technology are complementary. Blockchain provides distributed, secure processing and storage and a mechanism of trust, thus fostering decentralised cooperation, but is contract-agnostic. This knowledge is provided by ACTUS. It is composed of two parts, the data dictionary providing the format for the storage of contract-specific information and the algorithms that transform the contract-specific information (together with additional environmental information) into cash flows. The data format, as well as the algorithms, can be used on different platforms, among others on a blockchain. Thus, to create the infrastructure for future automated, digital banks, both BC/DLT and ACTUS are needed.
ACTUS seems to have some conceptual connection with other technologies, like HL7 adopted in Healthcare. Have you been inspired by a particular one during the definition of the project?
No, we haven’t. In fact, the roots of ACTUS go back some 25 years to the PhD thesis of Willi Brammertz. As co-founder of the Swiss company Iris, he then implemented these ideas in the risk management platform riskpro®, which today is part of Wolters Kluwer OneSumX solution. These roots in practice are a strength of ACTUS compared to other, more academic exercises.
I first heard of HL7 when the ACTUS project was already well underway. Meanwhile data standards exist everywhere. It’s important, however, to notice the intimate connection between the ACTUS data and algorithmic standard, which is essential but often overlooked. I’m not aware of something similar in HL7.
Do you think that ACTUS might also lead to the design of trading platforms of a new generation? For instance, by taking profit from a smarter data representation.
Yes, ACTUS is a general infrastructural element necessary for the automated straight-through processing of financial contracts from cradle to grave, and this includes trading. Typically, the design of trading platforms is much less complex than automating an entire bank because their scope is narrower, thus requiring only a subset of ACTUS’ capabilities. The strength of ACTUS is its power of integration, thus supporting the automation of whole banks and beyond, not only the trading division.
Some modern tools for analysing financial risk use network analysis (i.e. they consider the structure of connections among institutions). Do you think that this information can be embedded in your infrastructure?
It already is. Indeed, ACTUS also supports network analysis because it covers all relevant data and computations for financial contracts, which are the atomic elements that carry the financial interactions between different institutions. As such, they play a role similar to photons in physics which mediate the interaction between matter.
Specifically, ACTUS contains the counterparty information needed to reconstruct the network of financial interactions. Only the ownership control structure (e.g., blocking minority, etc.), is not covered by ACTUS, but this is addressed by the GLEIS initiative. Thus, ACTUS does not only support financial network analysis but, together with GLEIS, will give a boost to this approach.
The idea of adopting a common framework is, in principle, beneficial for many applications beyond finance. Do you have a new context in mind where a modified version of ACTUS might be useful?
ACTUS confines itself to financial contracts but the principles can be generalised. Indeed, Francis Gross, a colleague from ECB, is working on generalising the ACTUS principles well beyond the financial sphere so that, in principle, ACTUS could encompass all types of commercial contracts, including for assets, goods and services.
ACTUS is a cutting-edge digital solution for improving the management of financial contracts on a global scale. The ultimate goal is for ACTUS to become a standard framework in the finance industry.
- ARAMIS/Swiss Federation
- Alfred P. Sloan Foundation, (USA)
- School of Engineering, ZHAW
- Dr Willi Brammertz, Founder and CEO of Ariadne AG
- The Honorable Allan Mendelowitz, President of the ACTUS Financial Research Foundation and the ACTUS Users Association
- Prof Breymann’s Finance, Risk Management and Econometrics team, ZHAW
- Nils Bundi, ZHAW
- Prof Dr Kurt Stockinger, Institute of Applied Information Technology (InIT), ZHAW
- Dr Harald Stieber, European Commission
- Prof Petros Kavassalis, University of the Aegean, Greece
- Francis Gross, European Central Bank
- Johannes Micheler, European Central Bank
- The team of ODED, Hyderabad, India
- Jefferson Braswell, Tahoe Blue
Wolfgang Breymann comes from physics and turned to finance in 1996. In 2004 he joined ZHAW. He is one of the originators of ACTUS and board member of ACTUS Financial Research Foundation and ACTUS Users Association as well as founding member and Partner of Ariadne AG. His current R&D interests are focused on the automation of financial institutions.
Prof Dr Wolfgang Breymann
School of Engineering, Zurich University of Applied Sciences
Technikumstrasse 9 8401 Winterthur
T: +41 (0)58 934 78 14