In its bid to be an early adopter of AI, CalSTRS, the $358.4 billion pension fund for California’s teachers, has written a set of generative AI (GenAI) policies and guidelines, conducted an AI cost-benefit analysis and identified a pipeline of scalable use cases for the technology that all demonstrate business value.

CalSTRS is also well down the road on board and staff training efforts (staff will attend an AI bootcamp this September) to upskill the team in a sweeping change management effort.

Documents from a recent board meeting presented by April Wilcox, senior investment director, David Liao, director of enterprise IT solutions and innovation and Vaishali Dwarka, director of enterprise strategy management, laid out the next stage in a strategy where CalSTRS has committed to becoming a leader in the field.

As of August 2024, a total of 25 candidate AI use cases have been documented and analysed across more than ten business programs, six of which sit in investments.

Here, the technology will support predictive dashboards, cash flow forecasting and data augmentation and will help staff draft memos and knowledge share. It will be applied to portfolio and research intelligence, manager due diligence and facilitate the legal documentation behind the investment process, including reconciliation in private assets.

GenAI will be used to query data, generate summaries, gain insights into markets, ask questions and provide correlation analysis, they detailed. It will enhance risk management, optimise portfolio management and create operational efficiency where CalSTRS estimates the technology could increase efficiency by 85 per cent whereby processes that used to take 2.5 hours, take just 15 minutes.

In a joint presentation,  Accenture’s Ted Kwartler and Eyal Darmon detailed how integrating AI comes with important questions for institutional investors. CalSTRS must ascertain if its people, data and technology are ready for AI and investors also have to know where they will apply AI, and how to balance the value and risk of the technology.

Risks include the potential for unreliable outputs or the unauthorised disclosure of confidential information. Alongside security vulnerabilities like fraudulent attacks and disinformation, the technology also carries the risk of bias and will usher in changes in the workforce, which can exacerbate user mistrust of AI.

Integrating AI impacts investors’ risk exposure and shifts cost models. Decision-making models will be data-driven and CalSTRS will be dependent on ecosystem partnerships, they said. AI implementation also comes with multiple costs like tech infrastructure, data management and governance as well as the cost of running pilot projects, innovation labs and talent development.

The role of the board

In their presentation, Kwartler and Darmon also detailed the important role of the CalSTRS board in monitoring the rollout of AI.

For example, the board needs to stay informed about market trends which means monitoring the evolution of AI and how peer organisations are leveraging the tech. The board needs to encourage agility in AI decision-making to respond to industry shifts and have a firm grasp of the strategy. They also need to maintain an understanding of the AI program’s alignment with CalSTRS’ strategic goals and stay abreast of how AI investment is budgeted and its performance measured.

The consultants counselled on the importance of tying costs to mission-driven outcomes – suggesting starting small and scaling strategically. Quick wins are a good way to showcase upfront value which in turn builds momentum for long-term investments and will drive the transformation.

Using board notes that charted the evolution of AI, Kwartler and Darmon illustrated that Agentic AI, which acts on our behalf, planning and executing tasks end to end, is now capable of placing orders, scheduling repairs, or optimising workflows without needing a human to step in.

Alongside text (the most advanced domain, which has already passed medical, law, and business exams), AI is now able to write code for programmers and translate foreign languages.

AI also has multiple use cases outside the investment process including supporting payments to reduce repetitive interactions and freeing up staff time for other tasks.

Andrew Lo, finance professor at the MIT Sloan School of Management and quant investing pioneer, believes large language models (LLMs) can bridge the gap between fundamental and quantitative investing in a way that was unfathomable five or 10 years ago.   

While there are still a lot of “hallucinations” or misleading results in AI outputs, Lo said the finance industry is now at an “inflection point” in creating more effective ‘quantamental’ investment strategies which would bring together the best of both worlds.  

Lo, who founded quant alternatives asset manager AlphaSimplex in 1999, said the biggest difference between quant and fundamental strategies, or the obstacle standing between their integration, is fundamental strategy’s lack of scalability.  

“You don’t need a lot of quants to be able to manage billions, even hundreds of billions of dollars. It’s incredibly scalable,” Lo said during a presentation at the Machine Learning in Quantitative Finance Conference at Oxford University. 

“Fundamental analysis has its limits, because it just takes time and effort to bring all of these disparate pieces of information, much of which is not numerical,” he said. Taking his own healthcare investment venture as an example, Lo said the average fundamental analyst can only manage to follow 20 companies and come up with around five trades a year, and the only way to scale up is by hiring more fundamental analysts – until now.  

“It turns out that generative AI and specifically large language models, can now process both text and numerical information together,” Lo said. 

“It’s not just a bag of words that machine learning algorithms, when first applied to do natural language processing, tried to accomplish. It is much, much more powerful, and that is really a sea change which will allow fundamental analysis to be scaled.” 

While Lo acknowledged that no one has discovered how to specifically implement this scale-up yet, he is optimistic that a solution will come out in the near future.  

“Can quant and fundamental analysts live happily side by side and be able to be more productive than either by themselves? That’s what I mean by ‘quantamental’ analysis,” Lo said. 

But to make further progress on this integration, it’s essential that investors figure out how to conduct fair backtests of strategies involving LLMs. 

“I can run a backtest now of ChatGPT using data from 10 years ago. But the problem is ChatGPT has all the information over the last 10 years,” he explained. “I need to make ChatGPT dumber by somehow excising the last 10 years of knowledge, because otherwise it can just use the data for prices over the last 10 years, and yes, ChatGPT can perfectly predict future prices, given future prices.” 

“So here’s an idea [for the big AI firms]… create a financial product that allows you to license large language models by vintage – monthly large language models that are trained on data only as of that month and prior.  

“You can’t necessarily replay fundamental analysis over time, but you can certainly now run the quantitative part applied to fundamental data in a perfectly legitimate fashion.” 

Pattern spotter 

LLMs also have the potential to upend one of the most debated corners of investing, technical analysis, and lead to huge implications in fields like currency trading where pattern spotting is still prevalent, Lo said.  

Technical analysis refers to the method of examining past trading patterns of a security, commodity, currency or index to predict future price movements. Analysts decode the geometric shapes in graphs and charts that they surf through every day – ‘head and shoulders’, ‘double top’ and ‘double bottom’ and so forth – to identify signals and trading opportunities.   

Technical analysis’s reliability is often questioned and according to Lo, the very practice of it is looked down upon elsewhere in the industry with some referring to it as “voodoo finance”. It stands in contrast to fundamental analysts’ close examination of company balance sheets and macroeconomic factors and hinges its forecast on hints from historical market data, which are also highly open to interpretation.   

 “Someone once said… fundamental analysis is to technical analysis, like astronomy is to astrology,” Lo said. 

But at the heart of technical analysis is visual cues and pattern recognition, and Lo said AI has proven to be so much better at identifying them than humans. 

“They can actually identify patterns that humans cannot even begin to comprehend. We use double bottoms, head and shoulders, triangles, because we’re used to those shapes. How many shapes can large language models identify?” he said. 

“Large language models will become extremely good at trading foreign currencies among other asset classes. 

“Because in foreign currencies, if you’re trading, there’s really nothing else. Not a whole lot of changes between 10:03 and 10:04pm or am in terms of the trade deficit. So what are you going to do? You need to trade based upon something. This is an area where large language models can completely revolutionise the field.” 

The ultimate task 

In three years, Lo is optimistic that AI will be able to perform the most important mission in fund management: satisfying the fiduciary duty as defined by the law. 

Lo is currently conducting research around LLMs’ role as financial advisers and said it can already satisfy two out of three crucial requirements to provide sound financial advice: having expertise in the domain, and being customisable. The final criterion is developing trust and ethics, which Lo believes will be achievable in the foreseeable future.  

The boundary between human and AI is not as defined as we think, Lo said, and he pointed to prior academic efforts that aims to understand intelligence as a whole regardless of their forms, including Norbert Wiener’s exploration of cybernetics and John von Neumann’s posthumously published study, The Computer and the Brain, that discusses how the brain can be viewed as a computing machine. 

“My personal hero, Marvin Minsky… said ‘I don’t want to build a computer I can be proud of, I want to build a computer that can be proud of me’. 

“I believe that today we are on the verge of building computers that can be proud of us. That is both exhilarating and absolutely terrifying.” 

Singaporean sovereign wealth fund GIC boosted its US equities allocation in the year to March 2025 despite the expectation that high valuations could “provide a challenging backdrop for forward returns”.  

The equities allocation, which comprises public and private investments, now accounts for 51 per cent of the portfolio, up from 46 per cent in 2024, according to GIC’s annual report released on Friday.  

Fixed income makes up 26 per cent of the asset allocation, and real assets represent 23 per cent.  

The state investor now has an estimated $936 billion in assets under management, according to Global SWF, which also noted that GIC’s reporting has grown more opaque, now lacking specific asset class mix disclosures. 

The fund previously reported US exposure as a standalone category but now it’s grouped under ‘Americas’, which is the biggest geographical exposure (49 per cent) and up from 44 per cent in 2024.  

Despite its huge commitment, GIC questioned whether some companies in the US markets will be able to achieve the earnings growth implied by their high starting valuation.  

“A more challenging growth outlook weighs on their prospects. In this environment, our public equities teams are focusing on high-quality companies that can compound in value over the long term,” the annual report read.  

Since US President Donald Trump’s Liberation Day announcements, two schools of thought on the role of US assets in portfolios have emerged among allocators. Some are ready to slow the speed of capital commitment or set a limit of US exposure to manage risks, including Australia’s Future Fund and Canada’s IMCO.  

Others like Singapore’s Temasek stood firm on its position that it will keep funnelling capital into the US due to the depth and maturity of its markets, unless some extraordinary event – which it cannot foresee – forces it to pull back.   

But most allocators are no doubt carefully assessing the outlook. While uncertainties sparked by US policy changes around tariffs, immigration and government spending can lead to higher inflation and the probability of recession, GIC said “these worries are balanced by healthy private sector balance sheets, while in aggregate imbalances remain moderate”. 

“These will not just impact financial markets, but also hurt corporates who will need larger buffers to absorb these shocks. The result will be a less efficient and more costly trading system,” GIC said.  

GIC was established in 1981 to manage Singapore’s foreign reserves and it invests in international assets only. Its exposure elsewhere in the world includes 24 per cent Asia Pacific, 20 per cent in Europe, the Middle East and Africa and 7 per cent for the rest of the globe.  

In fixed income, “the low estimates of term premia indicate that medium-term around both inflation and deteriorating fiscal dynamics have not been fully priced in,” the fund said. 

It sees opportunities in real estate as valuations appear to be bottoming, while in infrastructure it is keeping an eye on themes such as digitisation, climate transition and emerging economy development.  

GIC doesn’t report on annual returns but the annualised USD nominal return for the past 20-year period came to 5.7 per cent as of March 2025.  

It employs a total portfolio approach with a reference portfolio that comprises 65 per cent global equities and 35 per cent global bonds, representing the fund’s risk tolerance. The reference portfolio’s 20-year annualised return is 6.2 per cent. 

GIC has kept the actual portfolio to a lower volatility than the reference portfolio over 20, 10, and five-year periods, which it attributed to asset diversification and “pre-emptive measures to lower portfolio risk” in recent years.  

Deputy group chief investment officer Bryan Yeo took over the top investment job from long-time CIO Jeffrey Jaensubhakij this April, who stepped down after almost a decade in the role. It has almost 2,400 employees across 11 offices around the globe.  

This article has been edited to clarify the rate of return and assets included in the fixed income asset class. 

The private markets team at the Teacher Retirement System of Texas (TRS), the $211 billion Austin-based pension fund, is increasingly concerned about the amount of retail money flowing into real estate and private equity.

Speaking during the investor’s July board meeting, Neil Randall and Grant Walker, who oversee TRS’ private equity and real estate allocations respectively, explained that the global assets under management for retail alternatives is around $4 trillion today but that figure is expected to increase in size to around $13 trillion by 2032.

It means retail investors could account for 22 per cent of the global alternative AUM by 2032 compared to 16 per cent today, eating into the slice of the pie available to institutional investors and bringing unwanted risks around liquidity, transparency and valuations to private assets.

Retail investors have become an increasingly rich source of fees and capital for GPs. Many have turned to retail investors because of capacity constraints within the institutional investor community in turn triggered by the lack of exits in private equity.

“We are above allocation to private equity so we are slowing the pace, but at same time the largest asset managers are creating products for retail and distribution to deploy products that are coming to market as we speak,” said Randall.

He added that the regulatory environment is also changing to put semi-liquid private market products into the hands of retail investors.

“Private equity firms with the largest platforms and strongest brands are expected to be the early winners.”

The TRS team is spending time “getting their arms around” the trend, conducting research internally and talking to other LPs, including industry body Institutional Limited Partners Association (ILPA).

Key concerns include transparency regarding the amount of retail capital raised alongside institutional capital in a particular fund. TRS wants to be able to work with GPs to devise a cap on the amount of retail money in the same fund they also invest in. Without this cap, retail dollars could suddenly swell the size of the fund and reduce the institutional weighting.

Other potential impacts from more retail capital flowing into private markets could be felt in LPs’ co-investment pipeline, eating into this valuable deal flow. It could also create potential conflicts between closed-end and semi-liquid funds.

It is just as big a concern in the real estate portfolio where TRS’ returns have been challenged by higher interest rates in recent years.

Substantial retail capital already flows into commercial real estate, one of the largest asset classes in the US. But Walker predicted that retail investors could sink up to $250 billion into commercial real estate annually.

“For GPs, this capital is more attractive because the fees are higher compared to what institutional investors like TRS pay. Institutional investors are also more engaged compared to individual investors who are more passive,” he said, noting that because GPs find this more attractive, some may begin to modify their business model by putting more focus on sourcing and managing capital from individuals or retail investors rather than institutions.

He said that TRS will try and mitigate the risk by leveraging its manager relationships and making much of its role as a large and long-term investor. The pension fund can act quickly and withstand market downturns without the need to withdraw money, he added

Trustees heard other key priorities in the private equity team.

These include reducing the allocation to 12 per cent in line with the new strategic asset allocation. Randall said progress is slow and steady. It will take another few years because of the risk of cutting back the allocation too quickly and suddenly being underweight.

The team are also continuing to move the portfolio from larger managers to mid-market managers with smaller fund sizes of around $3 billion. “It is going well, but we have more room to go,” he said, adding that this shift is being reinforced by retail flows into larger managers.

Another critical seam to strategy includes evaluating how AI will impact portfolio companies. “It’s important because every investment we make in private equity we will hold for 5-10 years within which time frame AI will reshape the economy. We have to think about this for every investment,” said Randall.

An external panel has recommended that Australia’s Sydney University minimise investments in defence and security-related industries within its A$770 million ($510 million) private asset portfolio rather than divest and book a A$67 million loss. 

The review was instigated by the University – which runs a A$3.4 billion portfolio comprised of donations and bequests – in mid-2024 following campus protests in response to the Middle East conflict. 

While the panel recommended the University sell its listed investments in aerospace companies, which generate revenue of about A$4.6 million, it deemed selling similar assets within its private equity pooled funds and fund-of-funds not feasible given the structure of those investments.

“It is not possible to impose exclusions on top tier private equity managers or upon private equity ‘fund of fund’ managers,” the report said.  

However, selling the entire portfolio on the secondaries market would invoke an estimated A$67 million discount on its current holding value, while the endowment would also miss out on future returns. 

“As also noted, the ‘opportunity cost’ of not investing in private funds (based on the current portfolio of A$770 million in investments) would be A$30 million per annum. This cost becomes exponentially greater, over time, given the effects of compounding returns.” 

Instead, the review recommended the University hold those private funds until maturity and identify any underlying investments that derive revenue from items listed on the Australian government’s Defence and Strategic Goods List (DSGL) Part One. It should disclose those investments annually and actively engage with private asset managers to stop any significant investment in those areas. 

The University of Sydney already excludes investments in cluster munitions, tobacco, fossil fuels and power generation companies that cannot demonstrate their transition to low carbon. 

The University is currently considering the findings of the report, which was overseen by ethics expert Simon Longstaff. 

global trend

The recommendation stands in contrast with a rising number of European pension funds which are loosening their ESG criteria to allow for greater investment in defence industries following pressure from the Trump administration. 

Denmark’s pension fund for academics, AkademikerPension – a strong proponent of applying an ESG lens across its portfolio – is the most recent fund to relax its rules for investments in the defence industry. 

“The security situation is currently worse than ever in recent times,” AkademikerPension CEO Jens Munch Holst said in a statement on the fund’s website in late June. 

“Russia continues its brutal war against Ukraine. And we see an increasingly clear picture of an imperialist Russia that clearly does not respect recognized borders and the right of other peoples to self-determination.  

“And at the same time, we can see that Europe, despite many talks with the American government, is very isolated when it comes to Ukraine, European security policy, the maintenance of democracy and relations with Russia. It is sad, and it calls for a new course in terms of investments in Europe’s defence.” 

It will now be able to invest in six European weapons manufacturers which have links to nuclear weapons: Airbus, Babcock International, Dassault Aviation, Leonardo, Saffron, and Thales. Three smaller companies have also been removed from its previous exclusion list (Serco Group, Groupe Reel and Ultra Electronics). However, the fund’s exclusion list still includes 46 weapons manufacturers. 

“We are not doing this to invest in nuclear weapons, but conversely, we do not want a small turnover from nuclear weapons-related activities to prevent us from supporting the capital construction of a European defence.” 

AkademikerPension’s statement was made on the same day that NATO leaders promised to boost annual defence spending to 5 per cent of their countries’ gross domestic product (GDP) by 2035. 

PFA Pension, the largest pension fund in Denmark, is another fund to recently allow investment in a select number of defence stocks. 

APG Asset Management which manages €552 billion ($650 billion) for Dutch pension fund Stichting Pensioenfonds ABP, is also considering increasing its existing $2.5 billion allocation to the defence sector, framing the argument in terms of security rather than buying weapons. 

Norway’s NOK 878 billion ($87 billion) Kommunal Landspensjonskasse (KLP), the fund for local government employees and healthcare workers, has just excluded US industrials group Oshkosh Corporation and Germany’s ThyssenKrupp for selling weapons including armoured personnel carriers, warships and submarines to the Israeli military.

“In June 2024, KLP learned of reports from the UN that several named companies were supplying weapons or equipment to the Israeli Defence Force (IDF) and that these weapons are being used in Gaza. On the basis of this information, KLP performed a thorough assessment of the companies and engaged in dialogue with them,” says Kiran Aziz, head of responsible investments at KLP Kapitalforvaltning, which she joined almost six years ago.

“Our conclusion is that the companies Oshkosh and ThyssenKrupp are contravening our responsible investment guidelines. We have therefore decided to exclude them from our investment universe.”

Although the combined value of the shares is minimal, KLP hopes to send a clear signal on the importance of human rights through divestment.

Last year, the investor sold its stake in Caterpillar due to the risk that the US company may be contributing to human rights abuses and violation of international law in the West Bank, and in 2022 it excluded 18 companies from its passive equity portfolio due to links with Israeli settlements in the occupied West Bank.

Still, KLP’s divestment from defence groups comes at a time when many European pension funds are examining their ESG policies to potentially invest more in listed defence companies. Geopolitical uncertainty and continued war in Europe, coupled with pledges from NATO leaders to increase defence spending to 5 per cent of their countries’ economic output by 2035, is fanning life into the sector.

KLP’s latest divestments do not reflect the investor’s preparedness to engage when it believes it can effect change. For example, although many investors have exited China, in recent years KLP has stepped up engagement with Chinese mining companies at risk of breaching its key concerns around labour rights and responsible extraction and mineral processing.

ESG is applied across the whole portfolio where strategy is focused on index portfolios offering broad market exposure and low cost, efficient asset management. KLP monitors and cajoles 7000 companies across 50 countries tracking MSCI and Barclays’ equity and bond indices, of which it currently excludes around 650.

“As an owner of 7000 companies globally you have quite a unique opportunity to set expectations and make sure companies have underlying economic activity that is responsible and sustainable,” she said. “We rely on publicly available information, and our expectations are levelled at boards and  management.”

KLP uses data providers to access information and get an indication of the level of risk. She is less focused on  ESG ratings but takes a keen interest in how a company is doing within its sector, using monitoring tools and working with other investors, stakeholders and civil society. She is in dialogue with around 300 companies annually.

The portfolio is divided between equity (35.1 per cent) short term bonds (26.5 per cent) long term bonds  (12.9 per cent) lending (11 per cent) property including Norwegian and international real estate funds (10.8 per cent)  and other financial assets (3.7 per cent)

KLP is the first company in Norway to have had its climate estimates approved according to the SBTi2 ’s new standard for financial institutions.

Aziz is a qualified lawyer who joined KLP with skills honed to argue and build a case following nine years at the International Commission for Jurists, the NGO that defends human rights and the rule of law. She is a board member of the Norwegian Refugee Council and the role takes her to refugee camps to meet people forced to flee which has informed her belief that investors have a responsibility to protect human rights.

She says the legal profession has taught her about the need to stand firm and persist and the need for courage to raise your voice when engagement is difficult.

“Many companies don’t want investors to interfere or tell them what to do, but investment is based on trust and companies should live up to certain standards.”