Dan Kemp has left Morningstar to set up Portfolio Thinking, an investment consultancy and research platform that will use AI to help advisers and other investment firms in their delivery of rigorous investment discipline and client communications.
As Portfolio Adviser revealed on Friday, Kemp departed Morningstar at the end of 2025 after over a decade with the group. He joined in 2014 as co-head investment consulting and portfolio management EMEA, and worked his way up to become chief research & investment officer of the global business of Morningstar in July 2023.
Prior to that he worked at Albemarle Street Partners, Saltus Partners and Williams de Broe, and spent the first eight years of his career was as a financial adviser.
After leaving Morningstar, Kemp has announced the set up of Portfolio Thinking and talks exclusively to Portfolio Adviser about the plans.
“I absolutely loved my time at Morningstar. It’s an extraordinary company with an incredible mission and I am very proud of the work that the teams I led there did. But there is nothing I enjoy more than working with advisers, helping them build and implement their investment propositions. There are a lot of challenges coming in in the UK, with consumer duty coming in and AI developing. These complications create an interesting time where I think advisers would benefit from support – I am going back to my roots and doing something I love.”
Investment engine
Portfolio Thinking will help advisers focus on long-term outcomes – portfolio construction, valuation discipline and behaviour through market cycles – and ensure the decisions surrounding them are robust as firms grow.
Kemp described it as a “scalable investment engine for modern advice firms” ranging from consultancy for investment committees to fully white-labelled investment propositions that integrate seamlessly into existing practices.
“We are in very noisy markets and it is easy to be focused on one thing or one event, and as you know advisers are outsourcing managed portfolios to providers such as Morningstar – some will use one, some will use a selection. That’s a very sensible way of running an investment proposition and for ensuring your clients have the best ongoing management.
“But the big enemy of an investor is not necessarily poor short-term performance, it’s the stress of being an investor. Whether that is the turmoil we saw in April last year or something to come in the future… people’s own cognitive weaknesses are one of the things that advisers have to work hard to address, challenge and manage.”
He added one of the most important roles of an adviser is to keep them on the path to their goals, and communicating easily at times of stress.
“My new role will be sitting with an adviser – in the background of an adviser’s business rather than the foreground like a traditional investment partner – and helping them assess their existing investment proposition, in the light of consumer duty to help them address any gaps that they have, and then provide that ongoing research support if they need that as well as ongoing governance support, which they’ll almost certainly need and then also – the most important thing – is the communication.”
For example, he pointed out if an adviser is using several portfolio managers, they may have several different market insights shared with them in times of stress, as well as client questions.
“How do you draw that together? How do you create something coherent? How do you create something that reflects you as an adviser practice helping those clients. Now, that’s something that is a challenge for advisers, not just in terms of doing the work, but the time it takes. And of course, there’s most call for that communication when everything else is at its most busy.
“This is where AI comes in.”
‘Supervised AI’
The investment engine offering by Portfolio Thinking provides ongoing fund oversight, risk monitoring, governance support and automated behavioural communication, using AI-enabled research capabilities. This will assist advisers by separating valuation mathematics from market narratives, probability-based equity valuations, fund analysis and investment theses designed to support robust product design and aid client communication.
“We are at this unique point in history with the emergence of AI. It’s going to completely transform investment research and advice, and it seems like such an incredible opportunity to try and help people through that, particularly advisers.
“A lot of people are asking – can AI do our investment research for us? But my view is that is the completely wrong way of approaching AI.
“We have perfectly good maths than can help us build out our research, and we have expert human judgement that can help in making some those investment decisions.
“But what AI does brilliantly is write and communicate because they are language models. That is what I have spent time doing, essentially building an AI engine that writes the way I would want it to write – to deal with clients and thinks about those behavioural issues, thinks about the core investment principles, and can produce that communication.
“When we marry that engine with an advisor’s chosen investment proposition, then my hope is that can really support the communication. I can help them build that investment proposition, help them manage that from a governance perspective and save them time. I think that is going to add value to the end client and comes back to consumer duty as well.”
Kemp is clear the firm is not going to be a regulated UK business and will not be managing portfolios. “However, we will be publishing model portfolios so we can be honest in terms of how we’re thinking about the world, and we will be publishing research.
See also: AI debt issuance is ‘transforming’ the corporate bond market
Governance
Kemp is also keen to emphasise the AI support is “supervised as there is a natural limit on what you can have it do unsupervised” when asked about potential risks highlighted with the governance implications around AI.
“One of the things AI is not very good at is facts and news as it hallucinates. It you ask it to tell you what happened in markets last week, it won’t do a good job of that. That’s where supervised AI comes into it and the governance of AI.
“It requires an expert to oversee it and that’s fine, the adviser will still be saving and huge amount of time on something that has been overseen by me, or in due course, someone in my team.”
As Kemp alluded to, he will be hiring further down the line but in the meantime he will be working with advisers to test the buiness model and ensure all the infrastructure is in place.
“I will be testing with advisers, and check the current thesis is correct and that we’ve identified something where there is a real need. You don’t want to be overconfident going into any new endeavour, you want to ensure it is thoroughly tested and you have the right service. We have had quite a lot of feedback already, and they are thinking about how we can help them specifically,
“Once you have identified that need, you can scale. We really want to provide as much help to advisers as we can,” Kemp added.
See also: Dan Kemp: ‘My clients are looking for answers!’
Portfolio Thinking will also be publishing a weekly research newsletter.
