Alexandre Lotfi of Bank of Singapore: Risk management, unstructured data and technology
Inside the Sell Side
Alexandre Lotfi
Global Chief Risk Officer | Bank of Singapore
Alexandre has over 20 years of banking experience in risk management across investment banks and private banks.
Climate change is one of the biggest risks we face today but it also presents several opportunities for us. This could be the last opportunity for us to change things and avoid the same mistakes made in the past.
Q. How has the risk function evolved over the years to cope with the digital future?
When I first started my career in the early 2000s, the focus of risk management was then on financial risks in areas like credit and market risks. Non-financial risks such as operational, business and conduct risks, however, have since evolved and are a bigger part of today’s private banking landscape. In fact, I spend 80% of my time looking at non-financial risks now as opposed to far less than 50% 10 years ago.
With non-financial risks growing in importance, we had to move away from traditional tools and ramp up the use of data to be more proactive in our risk management. The key challenge is having clean and accurate data to do so. Technology must complement systems to create solutions for risk managers to have tools to make the right decisions.
Despite our success in using data to create more accurate financial models and more effective stress testing, there are still areas to work on when it comes to managing non-financial risks. In the past, it was difficult to model or use traditional data to predict less ‘tangible’ risks. But today, with the use of alternative data and AI models, this area keeps getting better. For example, there are now models such as those used in weather forecasts and spatial data which can tell us how the crops will grow and the profitability for the next six to 12 months so that farmers can better manage their production and analysts predict more accurately the profitability of food companies.
Q. Bank of Singapore recently announced the implementation of a trade and communications surveillance tool using AI and data. Tell us more.
The solution is designed to connect disparate data across structured and unstructured data sources, with text analytics allowing the bank to delve deeper into conversation topics, sentiments and due diligence insights. We can create a holistic view of each customer profile without having our risk teams review sample accounts manually. Advanced analytics will also help the team uncover unknown behavioral patterns and relationships.
Given that our communication with clients is done via many channels and come in different forms – audio, video and text, it is imperative that such communications are captured and processed properly to ensure that the output is of a certain quality for analysis purposes. We certainly want to avoid the scenario of garbage in, garbage out.
Another area that we are mindful of is our people. When it comes to the success of any transformational projects, the people element is one of the most important aspects to look at. Some staff were initially hesitant to use a machine learning powered tool as they thought that their jobs might be at risk. But this will not happen because we still need risk managers to make decisions. The system will only give you signals when something is out of the ordinary. Someone needs to investigate whether it is right or wrong, and only human intelligence is able to do this very accurately.
Risk managers must be open to change and tap on technology to become more competent. Don’t be afraid of technology, embrace it.
Q. Sounds like getting risk management right involves being proficient in dealing with unstructured data.
When we started the trade and communications surveillance project, we collated unstructured, complex data in different shapes and forms. In dealing with unstructured data, we need to be forward-looking in capturing data and knowing what to do with it.
The first step was to clean the data and make sure it is accurate and timely so it could be used for decision making. Data needs to be well-defined and secure to use.
Voice data for example has to be transcribed to text before the systems can digest the data. For a start, we are focusing on three languages – English first, before we embark on Mandarin and Cantonese – amongst the 45 including dialects that are most commonly conversed between our clients and staff. While it is still not 100% accurate, but we are hopefully that it will be over time with ongoing tuning and also the deployment of machine learning – we give the system a conversation, look at the transcript, correct it and feed it back to the system to learn. The more you train the system the higher the accuracy is.
Q. We know you are an advocate for action on climate change, as its one of biggest risks that we face today. What can we all do more of?
Climate change is one of the biggest risks we face today but it also presents several opportunities for us. This could be the last opportunity for us to change things and avoid the same mistakes made in the past. Sustainability needs to be a key focus for risk management and the banking industry in general. We are doing quite a lot in that space to bring some real change. For example, we have incorporated ESG considerations in our lending ratio where clients are able to obtain a higher leverage when investing in mutual funds rated AAA or AA by MSCI. These funds are highly rated because they invest in companies that are leaders in tackling environmental, social and governance problems and have sustainability at the core of their business strategy. Indeed, this is a small initiative taken in isolation but there are more than 6 billion people around the globe and if everyone does a little bit, it is going to be a big step forward for the environment.
Q. What advice do you have for risk managers and how they can embrace technology?
Many are concerned that technology will take over their jobs. My role is to educate our staff, stakeholders and communities of the role of technology – to help us do things differently and better. Some jobs can be automated which is why we invest a lot in these fields, but we put our staff at the heart of any kind of transformation and technology will never 100% replace our staff.
Risk managers must be open to change and tap on technology to become more competent. Don’t be afraid of technology, embrace it. They must embrace this time of transformation through upskilling. To promote this, our staff are empowered to attend training without getting their managers’ prior approval. I have many staff who are currently learning coding – even if this is not a specific requirement for their role – but they are very passionate about it and passion must be the key driver in everything we do.