Country Head, Luxembourg
View bioCountry Head, Luxembourg
David Sarfas joined Intertrust Group in 2020. In 2022, David moved to Luxembourg where he took the role of County Head.
David is responsible for driving the growth of Intertrust Group in Luxembourg, the largest fund domicile in Europe. David is supporting the development of new funds and private capital-related offerings, working closely with clients and prospects alike to grow the business.
David brings with him over 25 years of experience in financial functions, serving large multinational corporations and private equity firms with extensive financial, accounting and transaction experience. Before joining Intertrust Group, David worked at MUFG Investor Services and has also previously held roles at EY and PwC.
David is bilingual French and English.
CloseOur report on technology for funds considers the benefits of robot process automation, machine learning, blockchain and other innovations for private capital funds in the future
In private capital fund administration, repetitive processes lend themselves to automation. The larger the fund, the more reliant it will be on automation because of the greater volumes of data and higher numbers of investors. Such funds will want to mutualise the costs of trading data and automate where they can.
That is not to say, however, that the way funds deploy technology needs to be automatic and recurring. Our major report, The Future of Fund Technology, digs deeper to consider what exactly is being automated.
A lot of funds rely on third-party administrators (TPAs) to provide the best possible technology to deal with their information.
Data around the volume of transactions is a large part of what general partners (GPs) or fund managers are dealing with. They are looking for a single stream of data, or so-called ‘golden source data’.
So it is no surprise that 57% of the 300 senior-level decision-makers in private capital firms we surveyed will rely on TPAs rather than in-house capabilities for data and technical requirements.
The automation of day-to-day administration must be aligned with the activity. So what is being automated and for what reason? The answer depends on the fund’s activity volume and strategy.
Funds pay fees to TPAs in the expectation that they are technologically best in class. Those funds not outsourcing may have very specific needs for automation choices, technology adoption and data requirements, leading them to conclude that a TPA isn’t an option.
Although funds reporting entirely manual processes are quite rare (8%), a further 22% said that the lion’s share of processes are manual. This is a substantial cohort of funds even if still in the minority.
The reason is that the smaller the fund, the more manual it is. Small funds may find the cost of investment in technology prohibitive. They may actually be using a TPA that can mutualise costs among many clients to deliver efficiencies that work for such a smaller fund.
In those funds using manual processes, a human being is required to carry out various actions. In a more technologically sophisticated fund these are done automatically – for example bank information is downloaded and allocated to the appropriate accounts by a system that can identify the relevant action.
Two technologies – robotic process automation (RPA) and machine learning (ML) – can make fund administration much more efficient.
Robotisation in the context of private capital means a real-world event generates a sequence of other events. For example, if a fund has a capital call of a certain amount, a calculation is done and allocated to all investors. An email is initiated with that information and sent to those investors.
ML or artificial intelligence (AI) offers very different functions. Here a programme is working in the background, analysing the human interactions associated with a particular task and identifying trends.
For example, in reconciliation, the machine stores all the actions to be carried out by the human who is identifying and reconciling the items. The machine begins automatically prioritising different elements to accelerate the process.
At Intertrust Group we have seen AI reduce reconciliation times from over five hours to under 30 minutes.
Indeed, AI has become a must-have for many clients, particularly larger funds. In our survey 38% of US respondents are ready to adopt ML, compared with 18% in Europe. This difference could be down to the regulatory environment creating fewer barriers in the US.
Two-thirds of respondents claim a very good or good understanding of blockchain. But interestingly that doesn’t translate into actual adoption or at any rate an appreciation of how to apply it.
The high level of stated understanding is surprising. It could be interpreted as a strong appreciation of how cost-saving and security aspects of blockchain might be leveraged for funds.
More accurate data gathered more cheaply is what distributed ledger technology (DLT) boils down to. With DLT, theoretically, a fund would need to do less due diligence because of inbuilt accuracy.
Blockchain could also have a huge impact if it were applied by auditors, with beneficial cost implications for funds.
Issues of interoperability and uncertainty over which platform will become standard offer some explanation for the relatively slow adoption, despite these cost and accuracy benefits.
We are now seeing DLT crop up a little in deal environments, though. And we expect to see this increase. There is little doubt that the gravitational pull of the technology will become more powerful over time.