Artificial intelligence has opened a new chapter in the technological transformation of independent fund managers. For Christophe Verbaere, however, it will only realise its full potential if it is underpinned by structured data and an organisation prepared to adapt its practices. For, beyond the technology itself, it is first and foremost the ability of companies and their staff to change their habits that will determine the pace of this transformation.
When it comes to artificial intelligence, what use case do you think currently creates the most value for an independent asset manager?
What we can expect from artificial intelligence today is, above all, increased efficiency. It enables a fund manager or operational teams to handle more portfolios and more clients without increasing resources. Doing more with what we already have, by freeing up time for high-value-added tasks rather than administrative or repetitive ones. It is hoped that, in the long term, AI will also help to improve performance, client relations and risk management, but there is still work to be done to demonstrate its capabilities and build confidence before this can be achieved. The real challenge, beyond the hype, is to bring about a lasting change in the way we work through a well-organised transition.
Why should an EAM begin their digital transformation with a genuine data strategy rather than with artificial intelligence?
Because the two are not separate. To unlock the power of AI, we must draw on internal knowledge, not just the external data that everyone has. We must draw on the organisation’s internal knowledge base, whether it consists of digital data, documents, site visit reports or communications. I have been convinced, long before AI came along, that a data-centric architecture is the key, even if it is more difficult to implement when it was not designed as such from the outset. When you can start with a blank page, you must place AI at the centre, which means that knowledge must be at the centre too.
For me, the two fundamental assets are the knowledge accumulated by the company and its staff. Everything else can change very quickly. A data strategy also helps to preserve this knowledge when employees leave – a long-standing challenge in banking and wealth management. Turning data into a structured asset that can be shared and utilised by the company therefore represents a genuine competitive advantage today.
You have led several technological transformation projects at NS Partners. What key lessons have you learnt from them?
Firstly, a great deal of patience. We are operating in complex environments, built up empirically, where technology has long been viewed merely as a support function. I’m fortunate at NS Partners to be a member of the executive committee, which isn’t always the case for those in charge of technology. Until technology is regarded as a strategic priority and a ‘business enabler’, it will be difficult to accelerate the pace of adoption.
As a second lesson, let’s talk specifically about adoption, which remains the most complicated issue. At company level, organisations prefer to see first what others are doing, and whether they succeed, before taking the plunge. We must therefore remain vigilant, ready to propose solutions. We must not overlook the need for support, which must be tailored to the needs and experience of each individual. Needs vary greatly; AI is a vast field that is constantly evolving and accelerating. Whatever the technology, people remain the most difficult factor to manage, and AI will not change that reality.
If you were to build the tech platform for a major independent asset manager from scratch today, how would you set it up?
I once had the opportunity to start from scratch, with the support of senior management. Technology is advancing so rapidly that it is impossible to predict what might change in six months’ time. The only certainty is that the company’s knowledge and skills constitute its strategic asset. Everything else – from tools and platforms to the organisation itself – can evolve.
In practical terms, you need to create a central hub that brings together and organises the company’s knowledge. The various technological building blocks then connect to this hub, without ever losing control over the data produced. There is no point in redeveloping the calculation engines or risk models provided by software vendors. On the other hand, the results must feed back into this common foundation. In this way, the company remains independent of its tools and can make full use of its data, particularly through artificial intelligence.
What will the next generation of PMS look like, and what should they look like?
Firstly, they will need to incorporate AI to augment the portfolio manager or relationship manager, without ever making decisions on their behalf, but by providing them with perspectives they might not have explored on their own. There is no need to recreate these models in-house. It is the role of software providers to integrate them directly into their solutions. It will be more effective to utilise the results and, where appropriate, develop one’s own agents as an overlay.
Furthermore, PMSs will need to offer better interoperability with counterparties. At present, too much time is spent connecting market data, custodian banks and external managers, which carries the risk of errors.
Finally, the PMS will no longer be able to operate independently of the client repository. Investment constraints and risk management will require us to have client information directly within the PMS, without compromising data confidentiality. At present, firms invest significant resources in creating interfaces between a PMS that manages portfolios and a central database containing client information, even though compliance checks rely precisely on this data. Once these elements are connected and enhanced by AI, we can envisage agents capable of optimising within constraints whilst taking into account the portfolio, the client’s profile and their environment. Let us not forget the difficulty of ensuring data security throughout this process.
Which roles, functions or skills are likely to evolve the most amongst independent asset managers over the coming years?
At NS Partners, we have chosen to invest in systems and artificial intelligence so that our staff are ‘augmented’ and to enable the existing structure to become more efficient. I do not believe in the idea, widely reported in the press, that a large proportion of jobs will disappear as a result of AI. This is a narrative we have seen with every technological disruption over the past century. The real challenge is one of adaptation, which remains more or less straightforward depending on individual profiles.
The central challenge remains adoption. We must first gain users’ trust by helping them reduce repetitive tasks and free up time for activities with higher added value. We must then gradually build their confidence in the results produced by the AI agents. Beyond AI, roles will also evolve with the development of digital tools for marketing and customer relations. In future, customers will expect to be able to ask questions directly to an agent capable of answering some of their queries.
Where will managers need to set their priorities?
Manual document review and data entry should no longer be necessary. These tasks are time-consuming, prone to errors and add no value. This is where AI can have the most immediate impact: by reading a document, extracting information and identifying inconsistencies. A due diligence process that used to take several hours can thus be reduced to a fraction of that time spent on verification and further investigation, with the rest of the time devoted to analysing and challenging the agent’s conclusions. We have implemented this internally, in particular to include more sources, improve quality and detect contradictions between documents.
Should independent fund managers launch their own technology initiatives straight away, or wait for specialised platforms to come onto the market to meet their needs?
A balance needs to be struck. Where it comes to elements that create genuine differentiation, asset managers must continue to invest. For everything else – that which clients expect but does not constitute a competitive advantage – it is better to let software providers gradually integrate AI and data into their platforms.
The key issue will then be connectivity between these different systems. Some platforms are already working on linking ‘digital twins’ together, an approach that could find new applications in finance. This may still seem forward-looking, but developments are already underway.
Ultimately, whichever system is chosen, the key remains data. Knowledge must become a corporate asset and be managed as such.