Identifying best practice in pension asset management is not a straightforward task. As much as asset allocators may want there to be a definitive answer, differences in size, mandate and resources between different pension funds means an investment approach that works for one may not work for others.    

At the recent Fiduciary Investors Symposium in Toronto, defined contribution and defined benefits pension funds from four different countries came together to discuss 30 years’ worth of governance, cost and performance insights collected by CEM Benchmarking, and it was clear that investors are still split on key issues such as the benefits of active management and scale. 

CEM data shows that active management has on average added 22 basis points of net value and 75 basis points of gross value per annum between 1992 and 2022 for asset owners. In public markets, the highest active risk and average one-year gross value added has been achieved in US small-cap equities strategies. 

It’s clear that there are benefits to active management in public markets, but South Carolina Retirement System Investment Commission is one fund that still decided to shift its entire public equities portfolio to passive, and chief executive officer Michael Hitchcock said it is not regretting the decision. 

“We had a significant amount of active management in the public equity portfolio – we had active managers, we had enhanced passive, we had GTAA [global tactical asset allocation], which was kind of a mix of public equity and bonds,” Hitchcock said. 

“And we really realised…out of humility that we’re really not that good at this. Because what we found is that we were underperforming the benchmark, at about the same amount that we were paying in fees. 

“We were able to shift to all passive, but we were able to do it in a way where through the structures that we have with our external partners [we can] pretty much lock in about a 30 to 40 basis point spread above the performance of the index, because of the efficiencies that these large managers have when it comes to securities lending, and the tax recaptured.” 

Hence, the fund is now outperforming by about the same amount that it was underperforming in the same part of the portfolio, Hitchcock said, and shifting to passive allowed the fund to utilise its resources better.  

“We decided that we really needed to put that energy into the private market asset classes, where we really felt like we would have the opportunity to outperform,” he said. 

However, UK’s Railpen doesn’t consider active management fees as a big issue. The fund has internalised two-thirds of its investment management. 

“On the public market side, we’ve got to about 50 per cent in equities; of that 50 per cent, we’ve got about 60 per cent in quantitative strategies and 40 per cent in fundamental portfolios,” said director of real assets and private markets Anna Rule. 

“Given I’ve seen their performance – yes, passive is cheap – but my fundamental team would argue that they’re both cheap. 

“They’ve been running that internal strategy for about the last eight years and have had about an outperformance versus ACWI of about 130 basis points.” 

The question of scale

When it comes to the scale discussion, the sentiment among funds is weighted heavily in favour of having greater scale, due to the potential access to deals, influence and cost structure benefits it may bring. CEM Benchmarking head of research Chris Flynn said the firm has not found evidence of diseconomies of scale so far. 

PSP Investments chief investment officer Eduard van Gelderen said being big has brought the fund many perks. 

“What I see happening in practice is that we are invited for specific deals, and there are only a handful of investors around the table [and] because we’re sitting at a table, we also dictate the terms of those deals,” he said.  

“Is there a limit to this? Well, we all know Norges Investment Bank, they’re so big that they have to basically invest passively all over the world. 

“I can only say for us there is a limitation on the public side, because I am not convinced that every active strategy we pursue actually has the capacity to do it. On the private side, at this point…there is no limit.” 

This view is echoed by Australian Retirement Trust, which gained massive scale in a short period of time from the merger of two Queensland-based pension funds, Sunsuper and QSuper. 

“Having spent the last couple of years on integration, we found that we really needed to focus on things that we’re going to be able to do well consistently – strategies that we won’t outgrow,” said the fund’s senior portfolio manager of investment strategy, Anne Blayney.  

“On the public market side, perhaps there will be some diseconomies of scale in that… as you get bigger and bigger, it’s going to be harder and harder to harness that alpha in some asset classes or some strategies.  

“But we haven’t we haven’t seen it as yet.” 

Financial services leaders need to step up in their understanding and application of AI, according to chief executive of Amundi Technology Ben Lucas, or risk being left behind.

Speaking at the Amundi World Investment Forum in Paris, Lucas said that the “opportunity is incredible, significant and immediate” but not enough financial services leaders were embracing AI.

“Your job will not be replaced by AI but it will be replaced by someone using AI,” he said. “At a company level your company won’t be replaced by AI but it will be replaced by companies that are using AI in an effective way. This needs to be embraced at a senior level, this is not someone else’s problem.”

Lucas said the financial services industry had done the hard work, including moving to the cloud, but needed to be “more bold and be doing more”.

“When you look beyond efficiency and look at effectiveness we think the upside could be exponential,” he told the 600+ crowd.

“We are not being bold enough and it is going to happen quicker than we think. We all need to embrace this, leadership all the way through the firm. Ultimately the biggest risk is doing nothing,” he said.

Also speaking at the event was Frederic Tardy, the head of financial services at Microsoft France, who said he had observed a lag in financial services use of AI.

“The challenge is how quickly you can learn gen AI, how you scale it, and that everyone in the company is learning it and grabbing it,” he said. “Senior execs need to be good at tech and be good at financial services.”

He urged executives to focus on getting the right team which means not just a lead engineer but also training the rest of the team.

“There needs to be practice. Senior leaders are speaking about AI but how many are using it every day? Ask yourself are you learning gen AI? Are you using it?”

Member of the House of Lords, Kay Swinburne, said financial services was uniquely positioned because it was already a highly regulated industry that was used to introducing new technology systems.

“We are in a good position where we can press the acceleration button in financial services,” she said, adding regulators also need to upskill.

“You need to view this as an area where you need to collaborate, not just choosing your partner in the sector carefully but you need to partner with your regulator and treat them as an advisor or partner too.”

Amundi’s chief operating officer, Guillaume Lesage, who is also founder of Amundi Technology said he was excited for the opportunities technology presented for the industry.

“AI enables people to concentrate on their core activities and use new data.”

And while there are risks, including ethics, transparency and cyber security, Lesage said good governance and security infrastructure are the antidote.

“This is going extremely fast and every month there is a new breakthrough. We are in a paradigm where it is risky to be too cautious,” he told the delegates.

Amundi has created an innovative hub with AI scientists embedded in the company to “create real business cases”, he said.

The firm’s own solution, the ALTO platform, now has more than 50 asset managers and €2.5 trillion on its platform.

Amundi is clearly positioning itself as a technology partner alongside its investment capabilities with Amundi Technology launched as a standalone business line in 2021 on the back of the internal front-to-back capabilities and skills, and the firm implementing software and transforming operational models as a result of its many mergers including Pioneer and Sabadell in Spain.

“We have been a partner on investment management for a long time now, we will be a partner for clients on technology,” Lesage said.

Lesage advised there were three key elements to success in technology: invest the right way; invest in people and bet on the right partners.

“Investment needs to be the right way, and not just everything and big investments everywhere. Invest in people. Success in innovation is not about budget it’s about having good people, entrepreneurs that take risks and are innovative. And bet on the right partners that will be there for a long time and will be in a dance.”

Amanda White was a guest of Amundi at the World Investment Forum.

The key to implementing a successful total portfolio approach (TPA) is not about creating complexities, but rather maintaining simplicity within the shared lexicon of an investment team, said Australian sovereign Future Fund and Canadian pension CPP Investments.   

At the Fiduciary Investors Symposium, head of total fund management at the C$632 billion ($460 billion) CPP Investments, Manroop Jhooty, said, TPA for the fund can sometimes mean sacrifices in sub-portfolio returns to ensure the bigger portfolio stays on track to meet mandate targets. 

It can be a somewhat “difficult” conversation to ask teams generating higher returns to hand back capital, Jhooty said, which is why building a common language and understanding in the investment team is important.  

“What the [TPA] dynamic is really about is relative value – how do you make the decisions around what assets you want to sell what, what you want to buy and why do you want to do that,” he said. 

“How we make those decisions is really having that common language to talk about what is the marginal impact to the portfolio [each decision has].” 

CPP’s TPA considers several risk-return factors such as economic growth, rates and credit spreads. When the fund rebalances its portfolio, Jhooty said it is essentially rebalancing to these macro factors and different asset classes provide exposure to the target factors in different ways. 

While those discussions can be difficult, “really what you need to be able to do is to have a common understanding of how all these pieces fit together,” he said. 

“That’s where TPA provides a great degree of grease for the wheels, so to speak, because you’re able to speak in a common language, and you’re able to view the portfolio through the lens of impact,” he said. 

Future Fund director of dynamic asset allocation Richard Cooney said while total portfolio frameworks should be rigorous, they also need to be “simple enough that people can have a common conversation”. 

“We look across asset class definitions and compare asset to asset, but sometimes that’s comparing it not an apple to an orange, but maybe like an apple to a loaf of bread,” he said. 

“You need ways to break them down and simplify the problem, so for us, that’s thinking about risk factors.” 

Cooney said getting the investment culture right is also critical. The A$223 billion ($147 billion) fund isn’t allowed internal management under its mandate, which is why its team has remained relatively small compared to the Canadian fund.  

“When people join the Future Fund, it feels different,” Cooney said. 

“It takes months to feel comfortable in the seat, and you have to get people to break down the silo mentality. 

“It’s not about how I’m doing against a benchmark. The benchmark doesn’t really mean anything. 

“You build in the mentality and the cultural awareness that it’s alright for me to sell this asset now or to not build at the same pace that I otherwise was going to, because I recognise…there’s a great opportunity over here, and if we didn’t slow the build right now, then we wouldn’t be able to lean in.” 

‘A degree of balance’ 

In practice, Jhooty said, CPP always sets out to create a balanced long-term portfolio which is meant to survive through multiple cycles and macroeconomic environments.  

“We’re looking for a degree of balance,” he said. 

“Where the discretion comes in from a tactical perspective is where are we today conditionally in the market…and how does that look different from what your long-term unconditional portfolio is. That’s where we begin to tweak some of those elements.” 

However, being “tactical” doesn’t mean being focused on “short-term blips in the market”, as Jhooty said CPP may consider how much growth exposure or inflation protection it wants. 

Jhooty said CPP considers each asset with certain degrees of substitutability when viewing the portfolio through TPA, which allows for some creative ways to rebalance the portfolio if needed.  

A good example is private equity, where many funds have been getting growth exposure in the last few years, and where there hasn’t been a lot of “activity on both sides of the equation”. 

“You didn’t see a lot of distributions coming back, nor did you see a lot of activity in terms of new capital being deployed,” he said. 

“Now, if you’re in a fixed asset allocation space, that creates a bit of a conundrum, where you’re potentially overweight, by virtue of the fact that activity is largely limited. Through TPA, which looks through, you’re able to more dynamically adjust and rebalance your portfolio.  

“Because what you really care about in that instance, is that you’re getting a little bit more growth exposure for your private equity portfolio than you were anticipating. But you don’t need to sell private equity, you can sell public equities, which also load on growth.” 

In Future Fund’s process, Cooney said the asset team doesn’t have a bucket to fill.  

“What we say to them is: you’re the experts, face your opportunity set, buy great assets… and work with people who are allowed to align to our total portfolio and return horizon,” he said. 

“We don’t really care what [assets] you call them, or where you where you find them, but there’s some guiderails at the top to make sure that we’re not aggregating risks in ways that we otherwise wouldn’t be comfortable with.” 

The fund also links performance-related benefits to total fund performance as a way to incentivise alignment with TPA thinking, Cooney said.  

“There’s no worry about if I do something, I’m going to cruel my ability to generate my own alpha – it’s the total portfolio return that matters.” 

David Bell, executive director of the Conexus Institute, reflects on the key themes of the Fiduciary Investors Symposium held at the University of Toronto last month including AI and its applications, the importance of governance for a strong pension model and common challenges faced by asset owners.

AI – fascinating developments, important core messages

The AI vertical stream (overviewing the technology, the investment opportunities, and the use opportunities for pension funds) was a feature of the symposium program.

Professor Ajay Agrawal’s keynote was my symposium highlight. By sharing his extensive AI domain knowledge through the filter of economic discipline, some important messages emerged:

  1. AI is all about prediction. Not all predictions made by AI will have equivalent levels of fidelity, something we need to be cognisant of. For instance, I am grounded by the difficulties of forecasting markets with any accuracy, so would be hesitant to rely strongly on AI. A nuanced trade-off between model insights and model transparency emerged through case studies. This needs to be balanced along with associated governance challenges.
  2. Focus on business outputs not inputs. Clarity on what you are trying to achieve with AI (business outputs) is a first step which will determine what AI inputs are required.
  3. Be aware of both point solutions and system solutions. Point solutions are those which affect part of a process, delivering productivity benefits. Here an investment application could be use of AI to review the data room associated with a private deal. System solutions are wholistic re-imaginings of how an output can be delivered, with transformational benefits. It is difficult to envisage system solution opportunities amongst asset owners due to…
  4. Friction of AI adoption will be greatest in regulated industry sectors. Regulated sectors will likely experience the greatest scrutiny around their use of AI. Pension funds and other asset owners generally fit in this category.
  5. The opportunity for clear market leadership (via the user feedback loop) explains the strong (and investment) focus by companies on AI. This presents an interesting investment dynamic for asset owners: a likely substantial dispersion between winners and losers, warranting consideration of a thematic opportunity but also an important implementation consideration (active vs passive).

Sustainability – Canada’s mature, balanced approach

As an Australian it was impressive to observe the degree to which climate and sustainability activities have been integrated into the investment processes of Canadian pension plans. Resolved, integrated objectives, mature positions on competing issues, acknowledgement of the broad trends in areas such as energy transition (along with associated risks), the setting of research-backed targets, and high-quality reporting were all strong features of Canadian funds.

In public markets there is a general acknowledgement of funds’ roles as universal owners, and the associated opportunities for active ownership and collaboration amongst with other asset owners. The opportunity for direct impact via private markets appears greater, due to an arguably longer-term focus (avoiding the short-term performance scrutiny experienced by public companies), and potential to have board representation.

The challenge remains that not every domestic investment will be financially appropriate. However, having integrated financial investment frameworks allows funds to make informed decisions, and positions them to explain clearly to governments the gap in the risk/return threshold. The opportunities appear strong for quality investors to work cooperatively with governments on impactful investment opportunities.

Addressing the capital requirements to finance the energy transition of developing economies appears less resolved and more difficult, yet many of the externalities will be borne by all.

Common challenges faced by asset owners

It was a privilege to study the ‘Canadian model’, from its origins (shared by Keith Ambachtsheer, Claude Lamoureux, Mark Wiseman and John Graham), and its foundations (independent governance, professional in-house investment management, scale, and extensive geographic and asset-class diversification).

However, the challenges faced by Canada’s system are broadly universal in nature. Two of these challenges relate to AI and sustainability, which have already been addressed. Other challenges of note:

  1. There are strong frictions to being a long-term investor. A non-exhaustive list includes weaknesses in governance models, performance reporting approaches, incentive structures, regulatory frameworks (for instance Canadian pension plans need to re-value every three years), and the role of media.
  2. Managing complexity and maintaining a degree of nimbleness. While Canadian pension plans tend to be at the complex end of the operating spectrum of asset owners (due to greater use of derivatives and leverage), the general trend of increasing scale and breadth of investment activity creates complexity which can be difficult to manage effectively (the example of risk systems was highlighted). An outcome can be a loss of portfolio nimbleness and associated opportunity cost. Exploring the benefits of simplification is an area likely to receive greater attention.
  3. Creating portfolio resilience was a focus, where I identified three levers for successful implementation: (1) TPA (total portfolio approach), an approach commonly adopted by asset owners operating in less benchmarked settings, (2) emphasising risk management which creates accounts for different regimes and, and (3) a cultural and governance-level preparedness to realise lower returns in the core (most expected) scenario as a trade-off for greater expected performance in other scenarios. Given the framework and risk tools already exist I consider (3) to be the threshold issue.
  4. Managing cultural challenges as asset owners evolve. Challenge could come from areas such as changes in scale, operations, investment opportunities, technology, and objectives. A common cultural challenge cited was private market teams being directed to slow down on new transactions. The importance of incentive structures aligning with culture was highlighted.

As always, the challenges are sizable, but I always take the view that challenge equals opportunity. Through that lens the opportunity for asset owners to improve outcomes, in multiple dimensions, is substantial.

Innovation in asset allocation and portfolio construction will come from creating the right structures to allow investment teams to flourish, the Top1000funds.com Fiduciary Investors Symposium heard. 

University of Toronto Professor of Finance Redouane Elkamhi told the symposium that he is “a big believer that structure drives everything”. 

“People realise that asset allocation matters way more than security selection. You [could] be like a genius at choosing stocks or bonds, but then … if you lose 40 per cent of the portfolio [that] is dramatic,” he said. 

“What have we noticed after that? We have noticed a rush toward doing some new structure thinking,” he said. 

“You have to give credit to thoughtful people trying to basically think outside the box and say we need to give possibility to management tilting portfolios and to make asset allocation value-add.” 

But he noted that “what you want people to do and what you design them to do are two different things”. 

“You want people to be able to do asset allocation tilting, and you also want them to do some security selection.  

“That’s really what you want them to do. You have to ask, do you [really] want them to do that? There’s nothing wrong, you may actually decide no, I don’t want them to do that. 

“It’s a choice. You agree it’s a choice? It doesn’t mean it’s wrong.” 

But Elkamhi said asking investment professionals to do one thing but putting them in a structure that doesn’t allow them to do that – or which encourages them to do something else – is counterproductive. He said it’s like taking a kid to a playground and then telling them they can’t play on the slide. 

“That’s a bad design,” he said. 

“You design the sandbox for which they can play, but then you let them play. Play means invest. Don’t design the sandbox, tell them you can play but you don’t really let them play. And I noticed this almost in every fund. 

“If you believe in asset allocation, hire the CIO that can do it. And if you hire a CIO that can do it, give them a structure that allows them to do it without screwing everybody else if they get it wrong. And that’s an important design. 

“There are some approaches people are thinking about it. It’s going quite well, hopefully things are moving. But that’s what I think [of] as new thinking in the pension industry in terms of structure.” 

Elkamhi said the key to innovation was bringing together two different types of thinking – one scientific and often largely theoretical, and one more like engineering, rooted in real-world solutions. 

“An engineer is making something work, but doesn’t have to find the law of physics,” he said.  

It’s the scientist’s job to define the laws of physics, “but you need an engineer to make something out of it”. 

“This is something I think missing a lot in the asset allocation – it is not missing in risk management, by the way, or pricing models – but in the asset management industry, you almost don’t see this bridging the gap easily between what I call an engineering-type of structure…and build-it-on-paper or the academic trying to find the best optimization, the best way to do things on the portfolio construction,” he said. 

Elkamhi said you “need a brain” to bring the two sides together, because an approach that is little more than crystal-ball gazing, or “you have somebody who just says ‘no, I have a feeling’, then the discussion, the communication can never happen”. 

Elkamhi said it’s important for pension fund boards to “create an environment that allows scientists to be very open minded about what they do. But [to solve] the problem you need to have an engineer that understands what they do”. 

“In my opinion, to flourish [you need] collaboration between the way of thinking about things as a scientist, and the way of thinking about this as an engineer.  

“It means don’t create silos, create collaboration. 

“You have to think of structure way more deeply. And I don’t think anybody yet solved that problem, because they just switch from one to one. And remember, there is no perfect solution for everyone. It’s just the consistency there is a problem between the incentive what the Board want [and] the structure you put in place.” 

Elkamhi that innovations in asset allocation also depend on being able to translate discoveries in academic research into practical real-world solutions. 

“One of the things I noticed personally through the years is that few research papers found adoptions in the asset allocation space. 

“You have almost like a tale of two stories,” he said. On the one hand, you have academics undertaking research largely for the sake of research, and on the other, you have professional investors trying to create portfolios that are both resilient and which meet the return targets needed to support pensions. 

Elkamhi said the role of a researcher in a university is to “get tenure and push the boundaries of research”. They do not need to concern themselves with the practical applications of what they discover. 

“That’s not part of the objective function of research,” he said. “But that doesn’t mean it’s not useful. 

“We need to spend some time to understand how to make them useful. And this is really what we’ve been thinking about.” 

Elkamhi said there is “very good innovation in the academic literature that actually can find very useful application in the industry”, it’s just a question of identifying it and then adopting it effectively. 

“You just need a way to bridge the gap between what have been developed in all the academic research.” 

Computing power has advanced to the point that the once-impractical process of reinforcement learning is now a viable tool for asset owners, the Top1000funds.com Fiduciary Investors Symposium has heard. 

Reinforcement learning trains software to make decisions by mimicking trial and error and is used in investment decision making to generate the best potential result. 

John Hull, Maple Financial chair in derivatives and risk management at the Joseph L. Rotman School of Management, told the symposium that reinforcement learning has several advantages and outperforms simpler modelling approaches. 

“It gives you the freedom to choose your objective function – it’s a danger with some of the simpler hedging strategies and so on that you’re just assuming good outcomes are as bad as bad outcomes,” he said. 

“You can choose your time horizon, tests indicate that it’s robust… and gives good results during stress periods and there’s a big saving in transaction costs. Why are we talking about it now? Well, because computers are now fast enough to make it a viable tool.” 

Hull said reinforcement learning techniques can reduce transaction costs by as much as 25 per cent compared with traditional hedging approaches. 

“It’s a way of generating a strategy for taking decisions in a changing environment – you’re not just taking one decision, but a sequence of decisions,” he said. 

“Perhaps you’re taking a decision today and then you take another decision tomorrow, and so on. Let’s suppose you’re interested in a strategy for investing in a certain stock and say what’s a good strategy for this stock – I think it’s going to work out okay, but it may not. What strategy should I use over the next three months. What do you do?” 

Hull said normally a stochastic process – which assesses different outcomes based on changing variables – would be used to assess a stock. 

“It’s uncertain how the stock price is going to evolve and you might use a mathematical stochastic process, you might use a historical data on the stock price behaviour, something like that. You have some model for how the stock price behaves,” Hull said. 

“Then your problem is defined by what we call states/actions/rewards.” 

Hull said the aim is quite simply to decide what action should be taken in each possible state to maximise the expected reward.  

“You’d say okay, we don’t know how this stock price is going to evolve but it will evolve in some way, and so there will be certain states we find ourselves in. We should take a certain action, and that’s what we’re trying determine, and there will be a certain reward,” Hull said.  

“In other words, you’ll make a profit or a loss. The way I think about it, it’s just sophisticated trial and error.” 

This means by starting off with having “no idea at all” about what a good action to take is and to then try different hypothetical outcomes. 

“It works well or it doesn’t work well, then you try a different action and so on and then eventually you come up with what seems to be the best action to take when a particular state is encountered,” Hull said. 

Hull said reinforcement learning traditionally is computationally expensive, takes a lot of computation time and is “data hungry”, but that’s not the case these days. 

“But fortunately, the other thing that’s happened that makes this a viable tool… is that we can now generate unlimited amounts of synthetic data that’s indistinguishable from historical data,” he said. 

“You collect some historical data… maybe a couple of thousand items of historical data [and] you can generate as much synthetic data as you want to that is indistinguishable from that historical data.” 

Hull said that while his experience has mostly been in applying reinforcement learning to the hedging of derivatives, he noted there’s many other areas where it can also be applied. 

“Because really it can be applied in any situation where the goal is to develop a strategy for achieving a particular objective in changing market,” he said. 

“There’s something out there that’s going to change in a way you don’t know, and you have to model that.” 

Financial Innovation Hub, or FinHub for short, carried out the research that Hull prestned to the symposium. 

Hull said one of the distinctive features of FinHub is that it’s not just academics within the Rotman School of Management that work on its projects, but also practitioners and the university’s engineering faculty. 

Reinforcement learning is just one of the projects FinHub has been working on, with Hull explaining the centre has also been doing work on natural language processing, amongst other initiatives. 

“We’ve worked with the Bank of Canada on monetary policy uncertainty,” Hull said. 

“We’ve done work on modelling volatility services and using natural language processing to forecast different market variables.”