IBSi Blogs

Machine learning and associated algorithms are making real waves not just in banks’ front offices but also in analysing and spotting trading opportunities in the stock markets.

Machine learning (ML) is being used to identify trading patterns, initially in historical trade and quote data. This may sound familiar to those who are cognisant of technical analysis and the end goal is indeed the same, that is, to find useful patterns in historical and even real time data that lead to decisions that may result in profitable trades either long or short.

According to Tom Finke, head of machine learning product management at software and data provider OneMarketData: “What’s different now is that the techniques have evolved in performing that analysis. In particular, we are able now to use machine learning algorithms to help improve some of the more historical sorts of algorithms that were used to try to detect patterns. When we train machine learning models, such as neural network models, they are able to find patterns that traditional analyses like regression analyses might not otherwise find. These sorts of algorithms are able to find patterns that mere humans are not able to find because of the extent of the vastness of the data that can be analysed.”

An arms race?

Of course, if every trader was to follow the same analytical signals they would all be doing the same thing at the same time. Finke admits that funds, brokers, trading firms, banks and asset managers are facing a “sort of arms race”! He added: If you don’t participate, there is a risk, you’ll indeed be left behind. It would be advisable for investment firms and investment funds that want to stay on top of things that they really should form teams to at least investigate how machine learning can help with their investment and trading decisions.”

But the machines are not completely taking over, well not yet. “It takes some human ingenuity and cleverness, to decide the parameters around those machine learning algorithms. For example, what is the most appropriate data set on which to build a machine learning model?”

Right now, the human element is still required. It takes a person to decide which ML algorithm should be used and what pool of data should be analysed. However, there are companies working on this with ML algorithms being developed with the aim of having them choose which are the best ML algorithms to use.

Watching the markets move

With algorithms being used to analyse trading patterns it should come as no surprise that one way this is being leveraged is in market surveillance. OneMarketData’s core product is a time series tick level database on which that company had built various vertical applications – one of the most popular being trade surveillance.

“We have that particular application being used by a major exchange in the US and by quite a few investment banks. They’re analysing order books; some historically and a few in real time to try to detect patterns that are nefarious such as spoofing  or layering [both forms of illicit manipulation in which a trader may attempt to deceive others regarding the true level of supply/demand for a given financial instrument]. Traditionally there are patterns that you can look for and detect to try to find these activities and now we are in the early stages of applying machine learning algorithms to that,” said Finke.

One thing that has changed in the financial markets in the last few decades is the sheer volume of data and number of trades. Finke noted that in the last week of February 2020 when the world’s financial markets became seriously ‘spooked’ by  concerns over the global Coronavirus (COVID-19) outbreak, the number of ticks during the volatility was such that “some of our customers were running out of memory in their memory databases”.

Seeking the right signal

ML and the appropriate algorithms are capable of analysing more than just price data. For example, Bloomberg now provide a machine-readable news feed that is tagged to make it easier for computer software to parse the text.

This parsed news may then be stored in the same way as trader quote ticks. Run it through a natural language processing algorithm to parse it into meaningful chunks (a technical term!) and then as stage two use another ML algorithm to decide whether there is enough information to provide a trading signal… and if there is, what should that signal be?

Such algorithms are being developed with the analysis of historical data but once the models are trained, they can be applied to real time streaming news data to try to generate real time trading signals. Sure-fire success in such endeavours is by no means guaranteed. “This still a hard problem. It’s hard for humans, it’s hard for anybody to pull meaningful market sentiment out of newsfeeds. We’re still in the early stages of having any sort of effective results, but that’s certainly not stopping people from making the attempt,” concluded Finke.

By Kiran Kumar, Co-Founder and Executive Director of Profinch Solutions.

Growth and relevance are quintessential business matters that keep organizations on the qui vive for opportunities to conduct business more efficiently and profitably while keeping step with changing times. The last few decades saw digitisation and technology emerging as this opportunity – starting off as a differentiator that set the leaders apart from the laggards to eventually becoming the only option available to stay relevant. The tech quarters are now abuzz with Quantum Computing – the nouveau arrive that promises to bring in the new wave of disruption.

Quantum Computing and Financial sector – What’s the fit?

Quantum Computing is a field which applies theories developed under quantum mechanics to solve problems. It entails the use of qubits to represent data as opposed to traditional binary units (0 and 1). Qubits are more flexible and allow for a combination of 0 and 1 simultaneously thus storing way more data than traditional bits wherein data must be either a 0 or a 1.

Quantum Computing’s enormous advantages over traditional computing stem from its conceptual design – the solution space of a quantum computer is orders of magnitude larger than traditional computers, even immensely powerful ones. The power of a quantum computer can be approximately doubled each time only one qubit is added. Relative to classical information processing, quantum computation holds the promise of highly efficient algorithms, providing exponential speedups in a multitude of processes.

Armed with these, Quantum Computing lends itself seamlessly to the financial sector since faster, more accurate, and more secure processing is at the core of how the industry needs to function.

Sample this – Google’s most advanced quantum computer named Sycamore could possibly solve a specific computational task that a traditional supercomputer takes 10,000 years to solve within 3 minutes. With that kind of speed and efficiency in tow, Quantum Computing is expected to produce breakthrough products and services likely to successfully solve very specific business problems. This could well usher in a new heyday, with financial sector holding the odds for being one of the most mightily favoured.

Delineating the Impact – what are the gains?

1. Enhance the efficiency of crucial operational processes in banking like

–  Client management, KYC processes, Client onboarding
–  Loan origination
–  Treasury management, trading and asset management

2. Revolutionise data security

Financial data encoded with quantum cryptography will be far more secure than other kinds of digital security. Such data cannot be hacked because the data in quantum states is perennially shapeshifting, i.e. constantly changing states and hence cannot be read. In fact, Quantum Computing has the potential to break even the most powerful security encryption of classical computers today. One of the examples to illustrate the use of quantum cryptography is known as a “quantum distributed key system” which promises secure digital communication that cannot be broken, even by a quantum computer itself. Banks such as ABN-AMRO are already starting to integrate this technology.

 3. Fraud detection

Quantum technology adeptly extends itself to fraud detection. As per a report in Feb 2019, financial institutions lose between USD 10 billion and 40 billion in revenue a year due to frauds and sub-optimal data management practices. Automation of fraud detection relies on recognizing patterns in data. Thanks to the qubit setup, the data modelling capabilities of quantum computers will prove superior in finding these patterns, performing classifications, and making predictions that are not possible today because of the challenges of complex data structures, thus averting fraud before it happens.

4. Customer targeting and service in banking

Classical computing is limited in its ability to create analytical models that can accurately and promptly cull insights from heaps of data available and target specific products at specific customers in near real-time. This greatly constrains the agility of response to rapidly evolving needs and behaviours of customers today. As per a study in 2019, 25% of small and medium sized financial institutions lose customers due to offerings that don’t prioritize customer experience. Quantum Computing can be quite the gamechanger for customer targeting and predictive modelling. It can also significantly enhance efficiency of critical frontal processes like customer onboarding which can sometimes take as long as 12 weeks to ensure due diligence. Use of quantum technology can turn around efficiencies thus enabling a far more superior and consistent customer experience.

5. Quantum data and transactions

Quantum technology’s ability to handle billions of transactions per second will be highly sought after by financial institutions consistently saddled with huge volumes of transactions. Quantum Computing reduces the likelihood of crashes and data loss. This will significantly accelerate the field of high-frequency trading.

Quantum computers will be able to mine colossal volumes of data almost instantaneously. This could enable the use of AI to make automated decisions using sets of pre-programmed rules.  AI is heavily reliant on large chunks of data to be able to learn. Given that Quantum Computing can handle that with incredible efficiency and speed, machines will quickly gather feedback that shortens their learning curve. Operations such as loan and mortgages can be automated, making them faster and efficient with seamless approvals and near zero delays.

6. Risk profiling

Financial services institutions are under increasing pressure to balance risk, hedge positions more effectively, and perform a wider range of stress tests to comply with regulatory requirements. With an ever-evolving regulatory climate, the complexity and cost of compliance is only expected to spiral in the coming years. Currently, Monte Carlo simulations are widely used to analyse the impact of risk and uncertainty in financial models but are highly limited by the scaling of the estimation error. In the face of more sophisticated risk-profiling demands and rising regulatory hurdles, the data-processing capabilities of quantum computers can improve the identification and management of risk and compliance.

7. Onward from here/ The shape of things to come

While the advantages run aplenty, Quantum Computing is still in the inceptive stages. A 2000 qubit quantum computer is expected only after 2025; beyond 2022, some aspects of Quantum Computing may start getting integrated with other cutting-edge technology of the day (such as AI and blockchain) to unravel amazing use cases in consumer experience, cybersecurity etc. The long and short of it is that, we stand at least five years away from Quantum Computing significantly impacting the financial services landscape. However, speculation abounds that Quantum Computing will mature at a velocity unseen by classical computing, and market developments and activities around it in the last couple of years endorse it. Reports say that financial bigwigs like Goldman Sachs, JP Morgan, CBA, Barclays, RBS, Allianz have already started investing in Quantum Computing technology.

The time is ripe for the penny to drop – for enterprises to start exploring investments in Quantum Computing. Those who adopt quantum early can seize major competitive advantages, including the potential to vault ahead of competition and become market leaders.

Successful compliance and risk management programmes within financial institutions depend on effective Know Your Customer (KYC) processes yet most have found these processes to be increasingly onerous, both in terms of time and cost. 

This is further complicated by an ever-changing and progressively stringent regulatory landscape. As a result, consortiums of banks, governments and vendors have explored the possibility of reducing costs and improving customer service through the establishment of industry-wide KYC processors or Utilities, with the intended aim of standardising KYC processes.

The latest region to look into adopting this approach is the Nordics, following a series of high-profile money laundering cases there. Leading banks DNB Bank, Danske Bank, Nordea Bank AB, Svenska Handelsbanken and Skandinaviska Enskilda Banken banded together in May last year to announce their intention to develop “an efficient, common, secure and cost-effective Nordic KYC infrastructure” called Nordic KYC Utility.

If designed and executed properly, the potential benefits of a shared infrastructure are clear to see, not least raising KYC compliance standards across the financial industry as a whole. However, there are challenges, as demonstrated by the Monetary Authority of Singapore shelving its plans for a centralised Utility late last year due to spiralling costs. Here, we look at what the objectives of a Utility and the key factors to be considered to ensure success.

Objectives of a shared KYC Utility

First and foremost, a Utility should provide benefits to ALL its stakeholders – from the participating banks and their customers to regulatory authorities.

For banks, the Utility should streamline KYC and customer onboarding, reduce costs and enhance KYC standards and auditability. Meanwhile, their end customers should benefit from a smoother customer onboarding journey and reduced friction.

From the perspective of the regulators, the primary objective of a successful Utility should be to raise confidence – among themselves and society as a whole – that banks are working to the highest KYC standards to more effectively combat financial crime, money laundering and terrorist financing.

Considerations for success when designing and implementing a Utility

The failure of the Singapore KYC Utility has highlighted areas of caution to be considered in other jurisdictions. In a report published by the Association of Banks in Singapore (ABS) after the project halted, it was revealed that “the overall margins at a systemic level did not allow for a viable business case in a projected term, and the proposed solution was going to cost more than the savings that banks would get out of it.” So, what can be learnt?

Approach the project with a thorough understanding of participant needs

It’s important to recognise that, despite all the participants operating under the same legislative and compliance regulations, their interpretations and – importantly – risk appetite will vary immensely. It is therefore crucial to establish a deep level of understanding of the needs of each participant in terms of existing KYC processes, risk methodologies, data and technology requirements.

The reality is one size never fits all. Participating institutions may vary significantly in their need to access various specific sources or adapt certain processes by jurisdiction or customer type based on their established compliance policies. This places extra demand on the Utility operator to select and deploy highly configurable best-of-breed technology, data and processes in the early stages.

Build design with flexibility at front of mind

The report highlights the importance of the core design, stating that “significant priority was given to design choices which represented a highly ambitious ideal” and that “more agility in governing the interaction between design and cost could have helped”. This emphasises the need for flexibility and adaptability, since initial requirements often evolve and so an agile approach and design thinking are essential to ensure the Utility truly delivers value based on the actual needs of the participating parties.

Ultimately, it was the cost of integrating an inflexible solution into a bank’s established compliance processes that proved to be the Singapore Utility’s primary downfall. The ABS report noted that the banks are “all at various stages of sophistication and evolution in terms of client data systems and KYC workflow systems”, and acknowledged that integration costs would account for over a third of the project costs. Therefore, ensuring there is flexibility in the design and technology used to build a Utility is critical in managing costs and ensuring participants realise maximum value today and into the future.

A Utility simply won’t work if its infrastructure is not future proof from advancements in regulation, technology or user experience expectations. For instance, here at encompass we recently launched global biometric identify verification (IDV), and this is the type of feature that Utility participants may well want to use in the future, so the infrastructure has to got to be able to adapt for this.

Are Utilities the way forward?

Given the amount of time and money financial firms currently spend on KYC and customer onboarding, it isn’t surprising that the Utility concept is gaining such traction. While there are certainly questions that remain to be answered – such as whether one Utility can realistically meet all the requirements of multiple, diverse institutions – the arguments in favour are certainly persuasive.

Industry pundits have debated the likely success of a regional KYC Utility and time will tell whether the proposed Nordic KYC Utility will achieve its desired outcomes. However, it is an encouraging move by the main Nordic banks to up the ante in the fight against financial crime, and rebuild their standing in the eyes of both regulators and customers. With a high level of engagement among key stakeholders, there shouldn’t be any reason why the project does not succeed, should it stay true to its primary objectives and understand that design thinking based on inherent flexibility is absolutely critical.

By Wayne Johnson, Co-Founder and CEO at encompass corporation

Research firm IDC is predicting banks worldwide will spend more than $4bn on Artificial Intelligence (AI) in 2018. If we factor in PwC’s Sizing the Prize report to understand the broader trend for global business, it seems AI could add a further $15.7 trillion to the global economy by 2030. Undoubtedly AI will lead to a significant change in the way banking operates, but let’s consider what that might look like.

By Dr Giles Nelson, CTO Financial Services, MarkLogic

Firstly, think about customer relationship management in retail banking. Most traditional banks still have a largely transactional relationship with their customers, providing deposit and payment services. But consider for a moment the information that a bank typically holds about an individual – their financial history gives a unique insight into a customer’s commitments, preferences and desires. By using AI techniques to analyse this treasure-trove of information, banks can deliver suggestions to tailored financial products and, indeed, other consumer products.

This kind of personalisation has begun with the introduction of chatbots in banking. AI is enhancing customer relationships by using natural language as a way in which customers can interact and ask questions and promises better customer satisfaction and lower call centre costs.

Fraud and anti-money laundering (AML) are also perennial issues. AI techniques, particularly using machine learning, are used today in these areas, and their use will only increase as models and databases of source information get more sophisticated. This is also a constantly evolving area as anti-fraud staff battle with bad actors who are also employing the latest technologies. With richer datasets and more advanced AI techniques, this will only get better, leading to less financial loss and less annoying false positives. With the right data of past and current transactions, the typical behaviour of customers can be learnt, and anomalies detected. Transactions can then be stopped, perhaps even before they have occurred, or confirmation from the customer requested before the transaction can proceed.

Last but not least is AI’s impact on trading technology. Much investment over the last 10-15 years has gone into making automated trading systems, whether trading equities, FX or derivatives, faster and more responsive to changes in the market. AI techniques, such as neural network machine learning systems, have also been used for some time. As AI tools and the data available to them become more sophisticated and richer, so these systems will get better.  Better at spotting opportunities to trade, and better at spotting the occasional examples of abusive behaviour.

Tackling a fractured data landscape

This is all seemingly beneficial and promises a lot, but here’s the rub. AI thrives on lots of data. To make AI useful, data from different parts of an organisation need to be accessible so the AI systems can use it. Data sitting in remote technology silos may be vital to a particular application, but if it isn’t easily accessible it may as well not exist. What’s more, that data has got to be well organised too – there is no area of technology where the aphorism ‘garbage in garbage out’ can be applied more strongly than with AI.

Furthermore, the data systems underpinning AI need to be agile enough to deal with new challenges quickly. Businesses cannot afford to spend months waiting for the right data to become available before launching new services – by then the competition will likely have an edge.

So, having the right data technology foundations is critical to delivering the process of AI, but a lot of banking organisations today don’t have this and are dealing today with a fractured data landscape.

The path to data-driven decision-making

Data silos are an undoubted issue together with the rigidity of most conventional data management systems. If financial organisations can go beyond this – delivering a holistic view of their data together with the agile data models that can evolve easily as business requirements change – then they can become truly data-driven. Competitive advantage will come from how smartly that data can be accessed and deployed.

More personalisation of retail services will occur, and banks will have the opportunity to strengthen their customer relationships and become more valued partners with end customers rather than just providing commoditised banking services. This will enable banks to provide services traditionally only targeted at the wealthy through private banking, to a much bigger segment of their customer base. Similarly, risk, fraud and money laundering should all ultimately reduced with the greater insights that AI can bring, making the whole financial system safer.

As with any new generation of technology, change can be both positive and negative, but one of the most scrutinised areas of disruption is the jobs market. With the introduction of AI, jobs will also change, and that’s the key point. One of the main purposes of any new technology is to make people more productive and to get ‘up the value stack’. This will need a willingness on behalf of people to embrace and, indeed shape, new AI-powered tools.

AI has the power to transform the banking sector, but only with the right data infrastructure. Banks should be acting now to ensure they have the right tools in place to make the most of the data at their disposal.

Kevin Day, CEO of HPD Software, a provider of technology that facilitates banks’ ability to provide asset-based finance, outlines the technologies banks should be considering to ensure that they are able to attract and retain the fastest growing SMEs as clients. They should embrace automation, offering integrated invoice finance platforms, upgrading mobile banking applications, introducing biometrics, and tapping into the vast potential of the blockchain.

In recent years, technology has fundamentally changed how the financial services industry operates. Fast-paced innovation in the FinTech sector has meant that SMEs should benefit from the cutting-edge technologies that are being developed when it comes to managing their everyday banking requirements. Yet many business customers are instead turning to more nimble digitally-driven platforms as traditional lenders have been slow to embrace such technology. This is beginning to change, however, as FinTech is starting to become integrated into the growth strategies of traditional lenders and other financial services providers that had previously taken a less tech-focused approach to the way they did business.

Mobile payments

Mobile payments may be old news for retail customers, but, in business banking it is now starting to catch up. In November 2018, Barclays became the UK’s first high street bank to launch a mobile invoicing application for its SME clients, which aims to reduce overhead costs and speeds up the payment cycle.

A recent report found that 80 per cent of businesses surveyed wanted time-saving measures to improve efficiency, with 56 per cent stating that these resources have a significant impact on their business and positively affect their decision to use what’s on offer by a financial institution. [1]Mainstream banks control 68 percent of the SME market share, so naturally technological innovation, like online and mobile banking, is playing a role in their continued dominance of the sector.

Biometrics

When it comes to business banking, online security is a concern that towers above the rest. Extending this functionality from a consumer platform into corporate banking has the potential to make payments and account management more seamless and secure. HSBC was the first to introduce biometric face, voice and fingerprint recognition for corporate clients in May 2018, and with its HSBCnet banking app, with single amounts of up to US$1 billion authorised on their platform, efficient, streamlined security is paramount. Biometrics has the potential to become significant for banks providing invoice factoring and other forms of asset finance, to authenticate the transfer of assets and to increase security around access to invoice finance platforms.

Blockchain

For such a tightly regulated industry, banks have historically been sceptical of whether blockchain technology is secure enough, but that attitude has shifted dramatically in very recent years. A recent Deloitte global survey finds that 43% of senior executives believe integrating and deploying blockchain to be one of their top-five strategic priorities.[2] Recent developments suggest blockchain technology is very quickly making the jump from niche application to mass market appeal, with growing consensus it will fundamentally change the way we think about asset financing and how banks and financial institutions operate.

We can see the benefit of blockchain in smart contracts – self-executing contracts on the blockchain, and in security – where transactions are immutable due to the decentralised nature of the distributed ledger. Among the areas where blockchain could be most beneficial for banks, is in their services for their asset based finance clients, as it can improve trade financing by keeping all stages of the process, including letters of credit and import/export authorisation, on the blockchain, which helps optimise settlement times, adds transparency, and lowers transaction costs.

Assed based finance tech priorities – automation and integrated platforms

The technology options available for banks are vast, and there are some which top the list, with digital, integrated, intelligent platforms being one of them. One of the areas poised to adopt technological innovation, where they can provide benefits for both banks and their business customers, is in facilitating asset-based finance.

ABF management software functionality and transaction streamlining have been transformed in recent years – the technology is now able to update collateral values as clients generate invoices in their day to day business, thus reducing the lead-time between raising an invoice and receiving funding. Banks should consider such platforms to automate and streamline data capture requirements and more efficiently deliver a seamless financing product to their business customers. Platforms such as our own HPD LendScape automates and streamlines data capture, offers real-time risk management, and provides insights and analytics into reporting.

As blockchain gains traction to support the business and trade transaction flows, there is an opportunity to connect ABF into the process; greater granularity and enhanced data quality can only help banks to provide enhanced funding into supply chains.

Outlook

Though mainstream banks initially appeared slow to react to the rapid changes, and adopt innovation with less agility then many alternative FinTech providers, they will likely remain their corporate clients’ main source of finance. However, clients will expect easy to use, fast and secure digitalised services as an integral part of the package or they will look elsewhere. Many major banks are acquiring FinTechs to take advantage of technical innovation and bring it to the mass market; we are starting to see the banking world rollout these technologies for the benefit of their business clients. If major banks want to stay ahead of the more nimble digital challengers, and retain and gain some of the fasted growing SMEs, they need to stay on top of the technology innovation game.

[1] https://www.pymnts.com/news/b2b-payments/2016/sme-mobile-banking-correlation-business-growth-performance/

[2] https://www.americanbanker.com/slideshow/the-many-ways-banks-are-using-blockchain

From the major European banks scaling back their trading units to the world’s largest investment managers slashing research spend by 40% – consolidation seems to be very much the theme of 2019.

By Daniel Carpenter, Head of Regulation at Meritsoft

Regardless of whether you sit on the sell or buy-side, here are four cost centres that financial institutions should take a hard look at if they are to harbour any hopes of seeking out much-needed efficiency savings from across the business.

1. CSDR: The Central Securities Depositories Regulation is one of the key regulations coming into effect in 2020. Though European, CSDR will affect all banks catering to European investors and is poised to be an operational nightmare if not handled properly. To comply with CSDR, banks will have to handle fails management, penalty of fails, buy-in process (sales settlement & reporting).

2. Preparing for more FTTs: Banks should be prepared for additional countries enacting financial transaction taxes in 2019 and the years to come. We’ve already seen talk of these taxes coming into effect in Spain and Germany, and rumours continue that there will be an EU-wide tax. Any banks that cater to ex-pats living abroad with US investments must be prepared to manage this slew of transaction taxes coming into play and consider what it means from an operational cost perspective.

3. Brokerage: Large banks are paying in excess of £100 million per year for brokers to facilitate transactions with counterparts across multiple desks. Banks can, of course, negotiate rates with their brokers. However, often the issue is that most have, until now, failed to find an accurate way to track and validate exactly how much they’re paying them per transaction and which rates are being used across desks, and across different units of the bank. Inadequate information, not being able to account for discounts, and a lack of comparability into how brokers charge for the same service, are all key factors behind failing to find an accurate way to measure and reduce costs.

4. Income management: While the primary focus of investment banks is making money, something not often considered enough is that banks also have to manage the process of giving investors money back in the form of claims and recovering Claims from counterparties. Claims arise in many forms, for example when coupons or dividends happen to come across mid flow between buying and selling. This means money can end up in the wrong accounts on more occasions than banks might consider, which can inadvertently become a significant cost center and risk mitigation aspect. With a significant amount of capital still tied up in old receivables, not to mention capital tied up in interest rate costs on outstanding receivables, banks need to seek out ways to track cash management and cash flows (receivables and payables) between their counterparties.

As 2019 begins to take shape, those that consider these four areas will be best placed to not only reduce operational overheads and funding costs, but to crucially handle mounting pressure from the boardroom to make efficiency savings.