AI has a role to play, but solving the missing middle may require a fundamentally different architecture for delivering wealth advice and investment access.
At The Wealth Mosaic’s AI in Wealth 2026: APAC event, held in Singapore on 24 September 2026, attendees considered how AI could unlock more scalable ways of serving one of the wealth management industry’s biggest but most challenging segments: the Asia-Pacific region’s growing mass-affluent segment.
Across the region, a growing number of investors sit between two established service models. At one end, digital platforms can provide efficient access to investment products for self-directed clients. At the other, traditional wealth management can provide highly personalised advice and servicing to clients whose assets justify the cost of a relationship manager (RM).
But between these two lies a “missing middle”: clients whose financial needs are sufficiently complex to require guidance, but whose economics do not support conventional, high-touch advice. The opportunity is not simply to put more products in front of mass-affluent investors, but to find a commercially sustainable way of providing the guidance and personalisation needed to help them make better investment decisions.
Laksh Gangwani, Chief Growth Officer at ViewTrade, a global investment and trading technology solutions provider, tackled the subject in a fireside chat with Patrick Donaldson, Founder at Mkt Dev APAC and Ambassador for The Wealth Mosaic.
Donaldson argued that assets under management are the wrong way to define the missing middle. What matters is the complexity of an individual’s or family’s situation, and therefore how much help they need. A client with relatively modest assets may have a business, cross-border ties or family obligations that call for real guidance, while someone wealthier with simpler needs may be well served by a self-directed platform.
Gangwani put it in commercial terms: “The middle is not a place. It is very different in different countries. It’s the amount of money that makes you unprofitable to service.”
Gangwani noted that investing involves a series of decisions: what to invest in, where to invest, which opportunity is appropriate, and how an investment fits into an individual’s circumstances. “All of those decisions are part of how you finally end up owning an asset. For those, you need somebody to guide you.”
The infrastructure problem behind the economics
“Is the missing middle fundamentally an economic problem, a service problem, or an infrastructure problem?” Donaldson asked.
“Why does the middle exist?” Gangwani answered. “It is an economic problem, it is a systems problem, it is an ecosystem problem. The whole thing is an infrastructure problem. And as soon as the industry realises that this is an infrastructure problem, and it needs to be solved like one, that’s when we are going to solve this.”
He argued that the industry consistently misdiagnoses transformation as a technology problem. “Every transformation is an infrastructure problem,” he said. “If you think it’s a technology problem, that’s a linear problem to solve. But no problem in this industry is linear.”
A seemingly straightforward initiative can quickly expose the dependencies underneath it, he said. Introducing fractional investing, for example, is not simply a matter of adding a capability to a digital front end. The order management system, back office, broker, and custodian all need to support fractional quantities. A project that begins as a front-end enhancement therefore becomes a much broader infrastructure transformation.
This has important implications for AI. Automating a process at the front end does not necessarily make the overall proposition more scalable if the trade subsequently encounters slow or manual processes elsewhere.
The productivity paradox
Donaldson pointed out that the mass-affluent segment also challenges the RM model. Even with hundreds of clients, an RM is incentivised to concentrate their attention on the most commercially valuable part of their client book.
Donaldson asked whether AI could genuinely help advisers serve more of those clients, or whether it would simply expose the limitations of the existing model.
Gangwani said AI offers a genuine opportunity, but not a magical transformation. Instead, it can remove friction from activities such as KYC, accessing suitability information, retrieving data, and generating proposals. The problem comes when that additional productivity runs into the rest of the system.
Donaldson highlighted a potential knock-on effect: much of the current AI conversation focuses on meeting preparation, CRM updates, client statements, and other front-office tasks. If those tools enable advisers to engage with more clients and generate more transactions, can the middle and back office cope with the resulting volume?
“Every front end problem will become your full transformation problem,” Gangwani said. Settlement cycles, different processes for equities, bonds, and structured products, manual execution, and other operational constraints can all become bottlenecks. “You will again have hit a point where they will be constrained by the slowest part of the workflow,” he said.
AI-driven productivity cannot be measured simply by how quickly an adviser can produce a recommendation, he argued: the rest of the workflow needs to be capable of supporting it.
From portfolios to financial lives
The conversation also moved beyond the question of how to make today’s model more efficient. Donaldson asked whether the portfolio itself should remain the organising principle of wealth management, or whether the industry should instead organise around the client’s broader financial life.
Donaldson’s own view was that the industry should move from episodic portfolio management to continuous financial management. Banks already hold much of the picture: income, spending, borrowing, savings and investments. Connected properly, that becomes a running view of a client’s financial health, closer to a fitness tracker than a quarterly review, and organised around the person and their objectives rather than the products they hold.
Gangwani said that wealth management has historically been built around compromises. Risk profiling and product matrices make personalised decisions economically manageable at scale, but can result in products being matched to predefined client characteristics, rather than starting with the individual’s complete circumstances. Although AI could change that equation by reducing the cost of handling complexity, that does not mean handing it the entire decision.
Donaldson asked where the boundary should sit between AI and human decision-making. He suggested the boundary should depend on the consequence of the decision. Much of the process is computational, such as assessing different scenarios, applying suitability criteria and presenting options within guardrails, but a recommendation with real downside is different. “That decision has to sit with the human,” Gangwani agreed.
That distinction matters for the missing middle: AI may allow advisers to serve more clients, but the adviser must still stand behind the output and take responsibility for what the client receives. For a client paying an advice fee, that professional responsibility cannot simply be transferred to a statistical model.
A middle that could get bigger
That leaves a question about where firms should actually begin solving the missing-middle problem. Gangwani rejected the idea that wealth managers can simply fix one horizontal layer at a time: the front end, then risk, then operations, then infrastructure. “If you ask me which layer would you solve first? That will not lead to results,” he said.
Instead, he argued for the “harder” but potentially more effective approach of solving the complete proposition vertically: take one specific service or customer need and address the technology, risk, operational, and regulatory requirements across the entire chain.
A workable vertical solution can then become a template for expansion, rather than another isolated technology project adding capability to an already fragmented architecture.
Gangwani said the industry needs to confront the problem now, because the missing middle is unlikely to disappear. Indeed, it may only get bigger: the supply of RMs is constrained while the number of people accumulating meaningful wealth is increasing. Advisers will continue to prioritise clients who generate the greatest economic return.
Technology alone will not cause those firms focused on serving other client segments to suddenly discover a profitable way of serving the mass-affluent segment, Gangwani said. The opportunity may instead belong to firms that make the mass-affluent segment their core proposition and build their technology, infrastructure, and operating model around it.
Although AI will be an important component of that model, it is not a solution to bolt onto existing wealth management processes. Instead, the industry should start asking what the process should look like if it were designed around the client from the beginning.
About The Wealth Mosaic
The Wealth Mosaic is a UK-headquartered online solution provider directory and knowledge resource, focused specifically on the wealth management industry.
For wealth managers, the buy side of our marketplace, The Wealth Mosaic is designed to enable discovery of key solutions, solution providers and knowledge resources by specific business needs.
For solution providers and vendors, the sell side of our marketplace, The Wealth Mosaic exists to support the positioning, exposure and business development needs of these firms in a more complex and demanding market.
The article was originally published in Wealth Mosaic’s Newsletter. You can read the article here.
