2

By Ken Schoff, Principal Partner (Sustainability); Manas Panda, Partner (Banking & Financial Services); and Raja Basu, Business Architect (Financial Markets), IBM

The timely and efficient Settlement of securities holds significant importance for the functioning of the global financial landscape. In recognition of this, regulatory measures have been introduced to address challenges associated with the settlement process. One such initiative is the final ruling regarding Exchange Act Rule 15c6-1(a) by the U.S. Securities and Exchange Commission (SEC), which emphasises the need to shorten the standard settlement cycle for certain securities transactions. By mitigating risks in clearance and settlement, this ruling aims to enhance the overall stability of the financial system and encourage technological innovation. Although subject to debate, the decision presents an opportunity for positive outcomes, particularly for those who prioritise global financial stability and the advancement of technology in the financial sector.

Technology’s role is of paramount importance to enable a seamless transition for any investment Bank. Transitioning to a T+1 (one business day after the trade date) settlement cycle in the United States requires investment banks to evaluate their technologies and infrastructures thoroughly. This evaluation should include an assessment of their existing systems’ capabilities in handling the increased volume and accelerated pace of settlements. Investment banks may need to consider upgrading or implementing new technologies, and one area that holds great potential is the integration of artificial intelligence (AI). AI can play a pivotal role in streamlining settlement processes, enhancing operational efficiencies and reducing the risks of errors.

Artificial intelligence can be leveraged to automate various tasks or improve the efficiency of existing processes in the entire trade lifecycle, albeit to various degrees at different stages in the process flow. For example, AI-powered algorithms can facilitate trade capture, confirmation and reconciliation, reducing manual intervention and the risk of human errors with expedited access to reliable digitised data throughout the process securely. Additionally, machine-learning (ML) algorithms can be employed to analyse historical trade data and pre-trade documents and identify patterns or anomalies indicating potential settlement risks. In toto, AI can improve operational efficiency across the trade lifecycle with enhanced experience levels.

The front and middle offices get a boost

Based on the observations of capital-market industry leaders, onboarding, trade booking and processing are some of the highest potential use cases of generative AI (GenAI) under consideration. AI-enabled trading platforms can not only assist in early-stage detection and resolution of risks but also fast-track the onboarding process by eliminating inefficiencies in the trade lifecycle, from booking to settlement to reconciliation to reporting. AI platforms and language models, combined with increased computing power, analyse both structured and unstructured data at a scale much broader than traditional statistical models. When effectively deployed, these capabilities can support straight-through processing by significantly reducing manual intervention and Enhance Operational Efficiency considerably. Simultaneously, AI can assist in better understanding clients and also design hyper-personalized products and services tailored to their needs.

A UK-based investment bank assigns significant importance to its AI initiatives, dedicating an entire subsidiary to realise its vision of becoming the country’s leading workforce in AI expertise. Within this framework, its in-house AI manages payments and trade decisions related to its short-term structured products.

We see a similar example in North America: A large investment bank in the United States has integrated AI technology to analyse loan agreements. As a key aspect of its AI strategy, the renowned contract-intelligence AI system has successfully saved a significant amount of time, totalling 360,000 hours, by efficiently interpreting and documenting contract clauses. This advancement has effectively freed operational staff from mundane tasks, allowing them to dedicate their efforts to more valuable and impactful responsibilities. Another prominent investment bank is exploring generative AI to assist investment advisors in scanning through piles of research reports and generate suggestions. This is expected to reduce search times significantly while enhancing the customer experience during trading hours.

Consistent monitoring of the regulatory space and fraud behaviour

In 2022, the SEC imposed US$6.4 billion in fines for non-compliance, while the number of fines issued by the FCA (Financial Conduct Authority) in the United Kingdom hit a record high with an increase of 160 percent over 2021. The rapid rise in the use of digital technology is making it increasingly difficult. Additionally, data-privacy regulations make the scenario more complex. This makes the perfect business case for implementing AI-enabled solutions. AI-based tools could help in avoiding regulatory intervention and penalties by red-flagging potentially fraudulent activities and settlement-failure risks.

At the same time, AI-based technologies, such as natural language processing (NLP) and robotic process automation (RPA), can aid in efficiently extracting and analysing accurate, relevant information for timely reporting, reducing compliance risks and enhancing the overall regulatory health of the organisation.

A UK-based global bank has leveraged AI algorithms, including machine-learning models, to bolster its fraud-detection capabilities in securities settlements. These algorithms can analyse large volumes of data and identify real-time suspicious patterns and potential instances of fraud. As self-learning models, they keep themselves updated with the changing dynamics and patterns.

Data management is key

To effectively incorporate AI into their technology infrastructures, investment banks should carefully assess their data-management capabilities. This includes evaluating data quality, accessibility and security measures to ensure that AI algorithms can leverage reliable and comprehensive data for decision-making. The more reliable the data, the better the outcome. Logical models work better when the dataset is broad, reliable and accurate. In this information-heavy world, when the level of interpretability and relevancy of data increases, the suitability of AI increases.

Building iron walls for security

The exponential growth in security breaches propelled by hybrid working models (systems that are outside of secure networks), deep fakes and others has raised threats and calls for robust guardrails. Additionally, banks should prioritise robust cybersecurity measures to safeguard sensitive client and transaction information, given the increased reliance on technology and data-driven processes.

By leveraging the potential of AI and upgrading their technology infrastructures, investment banks can enhance operational efficiency, mitigate risks and facilitate their smooth transitions to a T+1 settlement cycle. However, it is essential to strike a balance between the benefits of AI and the need for human oversight and control to maintain transparency, accountability and compliance within the settlement process.

Banks are rapidly exploring ways to incorporate AI and machine learning into their services, aiming to capitalise on the anticipated cost takeout and revenue opportunities of approximately US$34 billion to US$43 billion in the financial sector by 2025, as stated in a Goldman Sachs report. Similarly, as reported by Goldman Sachs, US companies pursuing or enabling AI have outperformed the broader stock market this year so far.

Another large European bank has incorporated AI technologies into its post-trade processes, encompassing security settlements and focusing on achieving cost savings. It utilises AI algorithms to automate trade matching, reconciliation and collateral management. Through the application of advanced analytics and machine learning, its objective is to enhance operational efficiency, mitigate settlement risks and expedite overall processing speed, thereby driving cost-effective outcomes.

In conclusion, artificial intelligence and machine learning have the potential to greatly facilitate the seamless transition to a T+1 settlement cycle in the US. By leveraging AI technologies, investment banks can improve operational efficiencies, develop stable straight-through processes, improve anomaly-detection capabilities by scanning through large volumes of data in real-time and mitigate risks, ensuring a smooth transition. However, it is always important to strike a balance between AI-human co-existence and cooperation to optimise benefits while maintaining transparency, accountability and compliance. Overall, integrating AI and machine learning can significantly contribute to a seamless transition to T+1 settlement in the US, enhancing efficiency, reducing risks and advancing the financial-market industry.

As Gal Krubiner, chief executive officer of Pagaya Technologies, said, “Efficiency in the world is coming through technology, and the more we can get these tools into the different parts of the system, the more we can get a result that is a better outcome for all of us as a society.”